Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
EcomScriptBench: A Multi-task Benchmark for E-commerce Script Planning via Step-wise Intention-Driven Product Association (2025.acl-long)
Copied to clipboard
Weiqi Wang, Limeng Cui, Xin Liu, Sreyashi Nag, Wenju Xu, Chen Luo, Sheikh Muhammad Sarwar, Yang Li, Hansu Gu, Hui Liu, Changlong Yu, Jiaxin Bai, Yifan Gao, Haiyang Zhang, Qi He, Shuiwang Ji, Yangqiu Song
| Challenge: | Goal-oriented script planning is used by humans to plan for typical activities . however, this capability remains underexplored due to several challenges . |
| Approach: | They propose a framework that enables product-enriched scripts by associating products with each step based on the semantic similarity between the actions and their purchase intentions. |
| Outcome: | The proposed framework can generate product-enriched scripts from 2.4 million scripts . human annotations are conducted to provide gold labels for a sampled subset . |
GraphNarrator: Generating Textual Explanations for Graph Neural Networks (2025.acl-long)
Copied to clipboard
| Challenge: | Graph representation learning has garnered significant attention due to its broad applications in various domains, such as recommendation systems and social network analysis. |
| Approach: | They propose to use a generative language model to map input-output pairs to explanations reflecting the model’s decision-making process to generate a model that generates pseudo-labels that capture the model's decisions from saliency-based explanations. |
| Outcome: | Extensive experiments show that GraphNarrator produces human-preferred explanations that are faithful, concise, and human-like. |
M-RewardBench: Evaluating Reward Models in Multilingual Settings (2025.acl-long)
Copied to clipboard
Srishti Gureja, Lester James Validad Miranda, Shayekh Bin Islam, Rishabh Maheshwary, Drishti Sharma, Gusti Triandi Winata, Nathan Lambert, Sebastian Ruder, Sara Hooker, Marzieh Fadaee
| Challenge: | Reward models (RMs) are primarily trained and evaluated in English and their capabilities in multilingual settings remain understudied. |
| Approach: | They construct a multilingual RM evaluation benchmark that tests the chat, safety, reasoning, and translation capabilities of RMs in 23 languages. |
| Outcome: | The proposed model performs better for high-resource languages and improves with translation quality. |
ELABORATION: A Comprehensive Benchmark on Human-LLM Competitive Programming (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on human-LLM competitive programming use scattered, application-specific human feedback. |
| Approach: | They propose a taxonomy of human feedback consolidating the entire programming process, which promotes fine-grained evaluation. |
| Outcome: | The proposed benchmark pinpoints strengths and weaknesses of existing methods and will be openly released. |
The Impossibility of Fair LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing frameworks for evaluating large language models do not extend to general-purpose AI contexts or are infeasible in practice. |
| Approach: | They analyze a variety of technical fairness frameworks to find inherent challenges . they find that each framework does not logically extend to the general-purpose AI context . |
| Outcome: | The proposed frameworks do not logically extend to the general-purpose AI context or are infeasible in practice due to large amounts of unstructured training data and potential combinations of human populations, use cases, and sensitive attributes. |
Intuitive Fine-Tuning: Towards Simplifying Alignment into a Single Process (2025.acl-long)
Copied to clipboard
| Challenge: | Supervised Fine-Tuning (SFT) and Preference Optimization (PO) are key processes for aligning Language Models with human preferences post pre-training. |
| Approach: | They propose to combine Supervised Fine-Tuning and Preference Optimization (PO) with two sub-processes defined at token level within the Markov Decision Process (MDP) |
| Outcome: | The proposed process performs comparably or even superiorly to SFT and some typical PO methods across several tasks, particularly those requires generation, reasoning, and fact-following abilities. |
Bias in Language Models: Beyond Trick Tests and Towards RUTEd Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Standard bias benchmarks are used for large language models to measure the association between social attributes and single-word outputs. |
| Approach: | They adapt three standard bias metrics of next-word prediction to measure gender-occupation bias and develop an analogous RUTEd evaluation in three contexts of real-world LLM use. |
| Outcome: | The proposed benchmarks are robust to lengthening model outputs via a more realistic user prompt in the domain of gender-occupation bias. |
Sliding Windows Are Not the End: Exploring Full Ranking with Long-Context Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for listwise passage ranking use sliding window approach, which is inefficient as it requires repetitive and serialized processing. |
| Approach: | They propose a listwise label construction approach and importance-aware learning objective for full ranking. |
| Outcome: | The proposed method outperforms existing methods in listwise ranking tasks. |
The Impact of Auxiliary Patient Data on Automated Chest X-Ray Report Generation and How to Incorporate It (2025.acl-long)
Copied to clipboard
| Challenge: | Traditionally, CXR report generation relies on data from a patient’s exam, overlooking valuable information from patient electronic health records. |
| Approach: | They propose to integrate patient data from ED records into multimodal language models that embed patient data into a language model. |
| Outcome: | The proposed model incorporates patient data from the MIMIC-CXR and MIMICIV-ED datasets to improve diagnostic accuracy and improves radiologist effectiveness. |
CLEME2.0: Towards Interpretable Evaluation by Disentangling Edits for Grammatical Error Correction (2025.acl-long)
Copied to clipboard
Jingheng Ye, Zishan Xu, Yinghui Li, Linlin Song, Qingyu Zhou, Hai-Tao Zheng, Ying Shen, Wenhao Jiang, Hong-Gee Kim, Ruitong Liu, Xin Su, Zifei Shan
| Challenge: | Existing studies have focused on the interpretability of Grammatical Error Correction (GEC) evaluation metrics, but the interpretabilty of these metrics has been neglected. |
| Approach: | They propose a reference-based metric that describes four aspects of GEC systems: hit-correction, wrong-corrections, under-correcties, and over-corrects. |
| Outcome: | The proposed metric reveals critical qualities and locates drawbacks of GEC systems. |
StrucText-Eval: Evaluating Large Language Model’s Reasoning Ability in Structure-Rich Text (2025.acl-long)
Copied to clipboard
| Challenge: | Structured data has been central to corporate data strategies for decades . however, with the advancement of large language models (LLMs), there has been a significant shift towards the effective utilization of unstructured data. |
| Approach: | They propose an automatic evaluation data generation method to assess LLMs’ reasoning capabilities on structure-rich text. |
| Outcome: | The proposed method supports 8 structured languages and 29 tasks, generating data with adjustable complexity through controllable nesting and structural width. |
Literature Meets Data: A Synergistic Approach to Hypothesis Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for hypothesis generation are theory-driven and data-driven, but they lack the computational power to complement each other. |
| Approach: | They develop a method that combines literature-based insights with data to perform LLM-powered hypothesis generation. |
| Outcome: | The proposed method outperforms baseline methods on five datasets and shows human accuracy improves on deception detection and AI generated content detection tasks. |
GAPO: Learning Preferential Prompt through Generative Adversarial Policy Optimization (2025.acl-long)
Copied to clipboard
Zhouhong Gu, Xingzhou Chen, Xiaoran Shi, Tao Wang, Suhang Zheng, Tianyu Li, Hongwei Feng, Yanghua Xiao
| Challenge: | Existing methods for achieving this require a limited understanding of constraints and can be hallucinating or brittle. |
| Approach: | They propose a framework that combines adversarial training dynamics with an encoder-only reward model to progressively learn and adapt to increasingly complex constraints. |
| Outcome: | Extensive experiments show that GAPO significantly outperforms existing methods like PPO, DPO, and KTO in fine-grained constraints. |
Tree-of-Evolution: Tree-Structured Instruction Evolution for Code Generation in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Data synthesis is a key research area in large language models (LLMs). |
| Approach: | They propose a framework that models code instruction synthesis process with a tree structure and optimization-driven evolution to alleviate constraints of unidirectional synthesis and randomness-driven generation. |
| Outcome: | The proposed framework outperforms open-weight code LLMs on five widely-used benchmarks. |
Delving into Multilingual Ethical Bias: The MSQAD with Statistical Hypothesis Tests for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Despite recent strides in large language models, studies have highlighted the existence of social biases within them . a recent study examined the ethical biase of LLMs concerning globally discussed topics . |
| Approach: | They propose to validate and compare ethical biases of large language models . they use news articles and socially sensitive questions to generate a data set . |
| Outcome: | The proposed dataset shows that ethical biases are widespread across languages and topics . the null hypothesis was rejected in most cases, suggesting biase arising from language differences. |
ReSCORE: Label-free Iterative Retriever Training for Multi-hop Question Answering with Relevance-Consistency Supervision (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-hop question answering requires reasoning across multiple documents to answer complex questions. |
| Approach: | They propose a method for training dense retrievers for multi-hop question answering . they leverage large language models to measure document-question relevance with answer consistency . their results lead to state-of-the-art Exact Match and F1 scores for MHQA . |
| Outcome: | Evaluated on three MHQA benchmarks, the proposed method improves retrieval performance . it leads to state-of-the-art Exact Match and F1 scores for the proposed technique . |
FACT-AUDIT: An Adaptive Multi-Agent Framework for Dynamic Fact-Checking Evaluation of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing fact-checking evaluation methods rely on static datasets and classification metrics, which fail to evaluate justification production and uncover the nuanced limitations of LLMs. |
| Approach: | They propose a framework that adaptively and dynamically assesses LLMs’ fact-checking capabilities by incorporating justification production alongside verdict prediction. |
| Outcome: | Experiments show that the framework differentiates among state-of-the-art LLMs, providing valuable insights into model strengths and limitations in model-centric fact-checking analysis. |
Statistical Deficiency for Task Inclusion Estimation (2025.acl-long)
Copied to clipboard
Loïc Fosse, Frederic Bechet, Benoit Favre, Géraldine Damnati, Gwénolé Lecorvé, Maxime Darrin, Philippe Formont, Pablo Piantanida
| Challenge: | Tasks are central in machine learning, as they are the most natural objects to assess the capabilities of current models. |
| Approach: | They propose a theoretically grounded setup to define the notion of task and compute the inclusion between two tasks from a statistical deficiency point of view. |
| Outcome: | The proposed model estimates the degree of inclusion between tasks on synthetic data and reconstructs the classic NLP pipeline. |
Towards Robust and Efficient Federated Low-Rank Adaptation with Heterogeneous Clients (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for federated fine-tuning for Large Language Models suffer from performance degradation at low ranks in heterogeneous data settings. |
| Approach: | They propose a low-rank adaptive model with Alternating freeze and Adaptive rank selection which reduces the number of uploaded parameters by 99.8% . |
| Outcome: | The proposed low-rank Adaptation maintains robustness even under extreme heterogeneity and low rank conditions while preserving communication efficiency. |
LLM-Powered Test Case Generation for Detecting Bugs in Plausible Programs (2025.acl-long)
Copied to clipboard
Kaibo Liu, Zhenpeng Chen, Yiyang Liu, Jie M. Zhang, Mark Harman, Yudong Han, Yun Ma, Yihong Dong, Ge Li, Gang Huang
| Challenge: | TrickCatcher generates test cases that pass existing tests yet contain bugs . a recent study found that tricky bugs are not detected by test suites . |
| Approach: | They propose an LLM-powered approach to generating test cases for uncovering bugs in plausible programs . they use a PUT and specification to generate program variants, an input generator and an Llm to construct test inputs . |
| Outcome: | The proposed approach achieves recall, precision, and F1 scores that are 1.80, 2.65, and 1.66 . trickCatcher generates program variants based on the program under test and its specification . |
Capture the Key in Reasoning to Enhance CoT Distillation Generalization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing distillation methods for Large Language Models (LLMs) focus on fine-tuning student SLMs on correct data, resulting in students struggling to learn the key instead of analyzing mistakes according to correct solutions. |
| Approach: | They propose a method that exposes key reasoning steps rather than simple fine-tuning students' CoTs data by using a set of prompts with similar reasoning paths but divergent conclusions. |
| Outcome: | The proposed method improves student SLMs' ability to learn key reasoning steps rather than fine-tuning them on teacher data. |
How to Enable Effective Cooperation Between Humans and NLP Models: A Survey of Principles, Formalizations, and Beyond (2025.acl-long)
Copied to clipboard
| Challenge: | Using large language models, intelligent models have evolved into autonomous agents . this paradigm has yielded remarkable progress in numerous NLP tasks in recent years . |
| Approach: | They present a review of human-model cooperation, exploring its principles, formalizations, and open challenges. |
| Outcome: | The proposed model-model cooperation paradigm has been a key focus of recent research . it is a novel paradigm that can be applied to a variety of tasks . |
Enhancing Hyperbole and Metaphor Detection with Their Bidirectional Dynamic Interaction and Emotion Knowledge (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for hyperbole and metaphor detection focus on superficial text features, ignoring the associations of hyperbola and metaphor . Existing frameworks focus on identifying superficial text, focusing on superficial features . |
| Approach: | They propose an emotion-guided hyperbole and metaphor detection framework based on bidirectional dynamic interaction. |
| Outcome: | The proposed framework outperforms baseline methods on four datasets. |
UniICL: An Efficient ICL Framework Unifying Compression, Selection, and Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve reasoning abilities of Large Language Models (LLMs) have limitations due to excessive growth in context length, causing large hardware burden. |
| Approach: | They propose a novel Unified ICL framework that unifies demonstration compression, demonstration selection, and final response generation. |
| Outcome: | The proposed framework unifies demonstration compression, demonstration selection, and final response generation. |
BelarusianGLUE: Towards a Natural Language Understanding Benchmark for Belarusian (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in NLP, such as large language models, have had groundbreaking impact on the field. |
| Approach: | They propose a benchmark for Belarusian, an East Slavic language, with 15K instances in five tasks: sentiment analysis, linguistic acceptability, word in context, Winograd schema challenge, textual entailment. |
| Outcome: | The proposed model underperforms on sentiment analysis, linguistic acceptability, word in context, Winograd schema challenge and textual entailment, but is competitive for linguistic acceptance. |
A Survey on Foundation Language Models for Single-cell Biology (2025.acl-long)
Copied to clipboard
| Challenge: | Existing single-cell foundation language models are based on pre-trained and large language models. |
| Approach: | They review the development of single-cell foundation language models . they discuss data tokenization strategies and pre-training paradigms . |
| Outcome: | The proposed models have shown remarkable performance in a variety of single-cell data analysis tasks. |
RuleArena: A Benchmark for Rule-Guided Reasoning with LLMs in Real-World Scenarios (2025.acl-long)
Copied to clipboard
| Challenge: | RuleArena assesses the ability of large language models (LLMs) to follow complex, real-world rules in reasoning. |
| Approach: | They propose a benchmark to evaluate the ability of large language models (LLMs) to follow complex, real-world rules in reasoning. |
| Outcome: | The proposed benchmark covers airline baggage fees, NBA transactions, and tax regulations. |
Extending LLM Context Window with Adaptive Grouped Positional Encoding: A Training-Free Method (2025.acl-long)
Copied to clipboard
| Challenge: | Existing long-context training data is scarce and requires substantial GPU resources for training. |
| Approach: | They propose a training-free plug-and-play method to enhance long-context understanding in existing large language models. |
| Outcome: | The proposed method outperforms existing LLMs on various tasks and surpasses baseline methods. |
Semantic Exploration with Adaptive Gating for Efficient Problem Solving with Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for multi-step reasoning suffer from inefficiency and redundancy . existing methods neglect the diversity of task difficulties leading to extensive searches even for easy tasks . |
| Approach: | They propose a method that explores reasoning paths with a gating mechanism that decides whether to conduct a tree search based on the confidence level of answers from a previous simple reasoning method. |
| Outcome: | The proposed method significantly improves accuracy by 4.3% on average while requiring only 31% of computational costs. |
HotelMatch-LLM: Joint Multi-Task Training of Small and Large Language Models for Efficient Multimodal Hotel Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | a novel multimodal dense retrieval model for the travel domain addresses limitations of traditional search engines. |
| Approach: | They propose a multimodal dense retrieval model that enables natural language property search . they propose combining a small language model and a large language model for embedding hotel data . |
| Outcome: | The proposed model outperforms state-of-the-art models on four diverse test sets . it is generalizable across LLM architectures and scalability for processing large image galleries . |
Can Multimodal Large Language Models Understand Spatial Relations? (2025.acl-long)
Copied to clipboard
| Challenge: | Spatial relation reasoning is a crucial task for multimodal large language models to understand the objective world. |
| Approach: | They propose a human-annotated spatial relation reasoning benchmark based on COCO2017 to improve MLLMs' spatial relation thinking. |
| Outcome: | The proposed benchmark achieves 48.14% accuracy, far below the human-level accuracy of 98.40%. |
S3 - Semantic Signal Separation (2025.acl-long)
Copied to clipboard
Márton Kardos, Jan Kostkan, Kenneth Enevoldsen, Arnault-Quentin Vermillet, Kristoffer Nielbo, Roberta Rocca
| Challenge: | Recent efforts to incorporate contextual representations into topic models have been shown to outperform classical topic models. |
| Approach: | They propose a theory-driven topic modeling approach that decomposes contextualized document embeddings into a Python package that implements S3 and all contextual baselines. |
| Outcome: | The proposed model is 4.5x faster than the BERTopic model and provides diverse and highly coherent topics with no preprocessing. |
TrimLLM: Progressive Layer Dropping for Domain-Specific LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | specialized language models do not show simultaneous memory saving and inference speedup at deployment time. |
| Approach: | They develop a layer-wise specialization technique that reduces the depth of LLMs by progressive layer dropping and compares it to other algorithms for inference. |
| Outcome: | The proposed model retains LLMs’ capacity in specific domains and achieves inference speedup irrespective of hardware and deep learning frameworks. |
JuStRank: Benchmarking LLM Judges for System Ranking (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work has focused on instance-based evaluation of LLM judges, where a judge is evaluated over a set of responses, or response pairs, while being agnostic to their source systems. |
| Approach: | They propose to validate the quality of the LLM judge itself by comparing system scores to a human-based ranking. |
| Outcome: | The proposed model fails to validate the quality of the judge itself, ignoring critical factors affecting system-level ranking, such as a judge’s positive or negative bias towards certain systems. |
Generating Diverse Training Samples for Relation Extraction with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models for Relation Extraction (RE) have good results on many benchmarks, but data scarcity is a common problem. |
| Approach: | They propose to use Large Language Models to generate training data for Relation Extraction . they propose to make LLMs produce dissimilar samples by direct instruction . |
| Outcome: | The proposed approach improves the diversity of training samples generated with LLMs while maintaining correctness. |
MultiSocial: Multilingual Benchmark of Machine-Generated Text Detection of Social-Media Texts (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for detecting social-media texts are limited to the English language and longer texts are not easily recognisable by humans. |
| Approach: | They propose to use a multilingual and multi-platform dataset to compare machine-generated text detection methods in the social-media domain to compare them to human-written texts. |
| Outcome: | The proposed dataset contains 472,097 texts, of which about 58k are human-written and approximately the same amount is generated by each of 7 multilingual LLMs. |
Efficient and Accurate Prompt Optimization: the Benefit of Memory in Exemplar-Guided Reflection (2025.acl-long)
Copied to clipboard
Cilin Yan, Jingyun Wang, Lin Zhang, Ruihui Zhao, Xiaopu Wu, Kai Xiong, Qingsong Liu, Guoliang Kang, Yangyang Kang
| Challenge: | Recent work utilizes feedbacks generated from erroneous cases to guide prompt optimization . previous methods rely on computational resources and powerful GPUs . |
| Approach: | They propose an automatic prompt engineering method that leverages feedbacks from erroneous cases to guide prompt optimization. |
| Outcome: | The proposed method surpasses state-of-the-art methods with less steps and lower computational resources. |
Evaluation of LLM Vulnerabilities to Being Misused for Personalized Disinformation Generation (2025.acl-long)
Copied to clipboard
Aneta Zugecova, Dominik Macko, Ivan Srba, Robert Moro, Jakub Kopál, Katarína Marcinčinová, Matúš Mesarčík
| Challenge: | Recent large language models generate disinformation news articles following predefined narratives . personalization and disinformation abilities of LLMs have not been studied . |
| Approach: | They evaluate the personalization and disinformation abilities of large language models . they find personalization reduces the safety-filter activations, thus effectively functioning as a jailbreak . |
| Outcome: | The proposed model generates disinformation news articles in english with the lowest quality of personalization. |
EscapeBench: Towards Advancing Creative Intelligence of Language Model Agents (2025.acl-long)
Copied to clipboard
Cheng Qian, Peixuan Han, Qinyu Luo, Bingxiang He, Xiusi Chen, Yuji Zhang, Hongyi Du, Jiarui Yao, Xiaocheng Yang, Denghui Zhang, Yunzhu Li, Heng Ji
| Challenge: | Existing language model agents excel in planning and reasoning, but lack creativity in unfamiliar environments. |
| Approach: | They propose a benchmark suite of room escape game environments to challenge agents with creative reasoning, unconventional tool use and iterative problem-solving to uncover implicit goals. |
| Outcome: | The proposed framework can perform with 40% fewer steps and hints and performs robustly across difficulty levels. |
BPP-Search: Enhancing Tree of Thought Reasoning for Mathematical Modeling Problem Solving (2025.acl-long)
Copied to clipboard
Teng Wang, Wing Yin Yu, Zhenqi He, Zehua Liu, HaileiGong HaileiGong, Han Wu, Xiongwei Han, Wei Shi, Ruifeng She, Fangzhou Zhu, Tao Zhong
| Challenge: | Existing datasets in operations research domain lack detailed annotations of the modeling process, focusing only on objective values. |
| Approach: | They propose an annotation-based tree-of-thought tree-based reasoning algorithm that integrates reinforcement learning into a tree- of-though. |
| Outcome: | The proposed algorithm outperforms state-of-the-art methods on StructuredOR, NL4OPT, and MAMO-ComplexLP datasets. |
LACA: Improving Cross-lingual Aspect-Based Sentiment Analysis with LLM Data Augmentation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to cross-lingual aspect-based sentiment analysis depend on translation tools. |
| Approach: | They propose a cross-lingual aspect-based sentiment analysis framework that leverages a large language model to generate pseudo-labelled data in target language. |
| Outcome: | The proposed approach outperforms translation-based approaches in six languages and five backbone models. |
Fusing Highly Specialized Language Models for Comprehensive Expertise (2025.acl-long)
Copied to clipboard
Ning Ding, Yulin Chen, Ganqu Cui, Xingtai Lv, Weilin Zhao, Kaiyan Zhang, Ruobing Xie, Bowen Zhou, Zhiyuan Liu, Maosong Sun
| Challenge: | Existing models that focus on language, programming code, and mathematical symbols are not able to achieve mastery of all three domains simultaneously. |
| Approach: | They propose to fuse highly-specialized models that are already sufficiently trained on different domains to achieve a highly-specific model. |
| Outcome: | The proposed model could achieve mastery of the three crucial domains simultaneously. |
HybGRAG: Hybrid Retrieval-Augmented Generation on Textual and Relational Knowledge Bases (2025.acl-long)
Copied to clipboard
Meng-Chieh Lee, Qi Zhu, Costas Mavromatis, Zhen Han, Soji Adeshina, Vassilis N. Ioannidis, Huzefa Rangwala, Christos Faloutsos
| Challenge: | Existing methods for retrieving information from a semi-structured knowledge base are struggling with hybrid questions. |
| Approach: | They propose a retrieval method that leverages both textual and relational information from a semi-structured knowledge base to answer user questions. |
| Outcome: | The proposed method surpasses all baselines on the STaRK benchmark and achieves significant performance gains. |
Re-ranking Using Large Language Models for Mitigating Exposure to Harmful Content on Social Media Platforms (2025.acl-long)
Copied to clipboard
| Challenge: | Social media platforms use machine learning and artificial intelligence to maximize user engagement, but can indirectly cause exposure to harmful content. |
| Approach: | They propose a re-ranking approach using Large Language Models to assess and rerank content sequences using large annotated data sets. |
| Outcome: | The proposed method significantly outperforms existing proprietary moderation methods on three datasets, three models and across three configurations. |
Aligning AI Research with the Needs of Clinical Coding Workflows: Eight Recommendations Based on US Data Analysis and Critical Review (2025.acl-long)
Copied to clipboard
| Challenge: | Clinical coding is labour-intensive and error-prone, which has motivated research towards full automation of the process. |
| Approach: | They propose to use AI to improve evaluation methods and propose new methods to assist clinical coders in their workflows. |
| Outcome: | The proposed methods can be improved and improved on existing methods and the existing ones to assist coders in their workflows. |
MIND: A Multi-agent Framework for Zero-shot Harmful Meme Detection (2025.acl-long)
Copied to clipboard
| Challenge: | a rapid expansion of memes on social media highlights the need for effective methods to detect harmful content. |
| Approach: | They propose a multi-agent framework for zero-shot harmful meme detection that does not rely on annotated data. |
| Outcome: | The proposed framework outperforms existing zero-shot approaches on three meme datasets. |
EvoWiki: Evaluating LLMs on Evolving Knowledge (2025.acl-long)
Copied to clipboard
Wei Tang, Yixin Cao, Yang Deng, Jiahao Ying, Bo Wang, Yizhe Yang, Yuyue Zhao, Qi Zhang, Xuanjing Huang, Yu-Gang Jiang, Yong Liao
| Challenge: | Existing knowledge evolution benchmarks are static and fail to capture the evolving nature of LLMs and knowledge. |
| Approach: | They propose an evolving dataset that categorizes information into stable, evolved, and uncharted states. |
| Outcome: | The proposed dataset is auto-updatable and enables evaluation of continuously changing knowledge and newly released LLMs. |
Rethinking Repetition Problems of LLMs in Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have focused on content repetition, but structural repetition is a more prevalent problem in code generation. |
| Approach: | They propose a decoding approach that eliminates repetition problems in code generation by identifying grammar rules and strategically decaying the likelihood of critical tokens that contribute to repetitions. |
| Outcome: | The proposed approach outperforms baselines and humanEval benchmarks on CodeRepetEval dataset and MBPP benchmarks, effectively reducing repetitions and enhancing the quality of generated code. |
PunchBench: Benchmarking MLLMs in Multimodal Punchline Comprehension (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks on punchline comprehension suffer from language shortcuts that allow models to rely on text, lack of question diversity, and narrow focus on a specific domain of multimodal content. |
| Approach: | They propose a multimodal punchline comprehension benchmark to assess models' ability to comprehend punchlines. |
| Outcome: | The proposed model surpasses in-context learning and chain-of-thought in punchline comprehension. |
ProcessBench: Identifying Process Errors in Mathematical Reasoning (2025.acl-long)
Copied to clipboard
Chujie Zheng, Zhenru Zhang, Beichen Zhang, Runji Lin, Keming Lu, Bowen Yu, Dayiheng Liu, Jingren Zhou, Junyang Lin
| Challenge: | Existing models fail to generalize to more challenging math problems, authors say . existing benchmarks related to assessing language models' reasoning process are limited . |
| Approach: | They propose a tool to measure language models' ability to identify erroneous steps in reasoning . they use two types of models: process reward models and critic models . |
| Outcome: | The proposed model outperforms existing models in evaluating language models' reasoning process . the best open-source model has demonstrated the critique capability competitive with the proprietary model . |
Model Extrapolation Expedites Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve LLM alignment training require expensive computational resources. |
| Approach: | They propose a model extrapolation method to expedite LLMs’ alignment with human preferences by amplifying parameter changes based on a first-order approximation without any additional training overhead. |
| Outcome: | The proposed method outperforms a fully-trained model on leading benchmarks and significantly outperformed open-source models. |
ATLANTIS: Weak-to-Strong Learning via Importance Sampling (2025.acl-long)
Copied to clipboard
| Challenge: | ATLANTIS is a new technique to improve the performance of large language models. |
| Approach: | They propose a new technique to bridge the gap between the distribution of current datasets and the real-world data distribution by using importance sampling. |
| Outcome: | The proposed technique can bring consistent and significant improvements to models’ performance and can be flexibly transferred among models with different structures. |
MPVStance: Mitigating Hallucinations in Stance Detection with Multi-Perspective Verification (2025.acl-long)
Copied to clipboard
| Challenge: | despite advances in large language models, challenges persist due to hallucination-models generating inaccurate content. |
| Approach: | They propose a framework that integrates multi-perspective verification with Retrieval-Augmented Generation to address these challenges. |
| Outcome: | The proposed method outperforms existing models on the SemEval-2016 and VAST datasets. |
Personality-Guided Code Generation Using Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have shown that personality-guided code generation improves software development outcomes when individuals are assigned tasks that match their personality types. |
| Approach: | They evaluate how emulating personality traits appropriate to the coding tasks affects LLM performance by using seven widely adopted LLMs. |
| Outcome: | The proposed approach improves pass rates in 23 out of 28 LLM-dataset combinations, while emulating personality traits can be easily integrated with other prompting strategies to further boost performance. |
PsyDT: Using LLMs to Construct the Digital Twin of Psychological Counselor with Personalized Counseling Style for Psychological Counseling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing mental health LLMs do not consider the fact that different psychological counselors exhibit different personal styles. |
| Approach: | They propose a framework that uses LLMs to construct the digital twin of psychological counselor with personalized counseling style. |
| Outcome: | The proposed framework can synthesize multi-turn dialogues that closely resemble real-world counseling cases and demonstrate better performance compared to baselines. |
BIPro: Zero-shot Chinese Poem Generation via Block Inverse Prompting Constrained Generation Framework (2025.acl-long)
Copied to clipboard
| Challenge: | generative pre-trained models face challenges on constrained writing tasks like poem generation . brian mccartney: BIPro improves the zero-shot generation quality on constricted writing tasks . |
| Approach: | They propose a framework that leverages two block inverse prompting methods to improve the quality of constrained writing tasks. |
| Outcome: | BIPro significantly improves the quality of Chinese poem generation without priming or training. |
LongDocURL: a Comprehensive Multimodal Long Document Benchmark Integrating Understanding, Reasoning, and Locating (2025.acl-long)
Copied to clipboard
Chao Deng, Jiale Yuan, Pi Bu, Peijie Wang, Zhong-Zhi Li, Jian Xu, Xiao-Hui Li, Yuan Gao, Jun Song, Bo Zheng, Cheng-Lin Liu
| Challenge: | Existing document understanding benchmarks only handle a small number of pages . existing models are limited to handling only a limited number of documents . |
| Approach: | They propose a long document understanding benchmark that integrates three primary tasks and 20 sub-tasks based on different primary tasks. |
| Outcome: | The proposed model outperforms existing benchmarks on open-source and closed-source models . the model outpersforms other models on more than 33,000 pages of documents . |
ObfusLM: Privacy-preserving Language Model Service against Embedding Inversion Attacks (2025.acl-long)
Copied to clipboard
Yu Lin, Ruining Yang, Yunlong Mao, Qizhi Zhang, Jue Hong, Quanwei Cai, Ye Wu, Huiqi Liu, Zhiyu Chen, Bing Duan, Sheng Zhong
| Challenge: | Recent studies show that obfuscation techniques for MLaaS are susceptible to embedding inversion attacks (EIAs). |
| Approach: | They propose a model obfuscation framework that protects client inputs from embedding inversion attacks by obliviously obbing models. |
| Outcome: | The proposed framework outperforms existing works in utility by 10% with a nearly 80% resistance rate against embedding inversion attacks. |
Interlocking-free Selective Rationalization Through Genetic-based Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to selective rationalization suffer from interlocking, a phenomenon known as interlock. |
| Approach: | They propose a genetically-based disjoint training architecture for selective rationalization that avoids interlocking by performing genetic global search. |
| Outcome: | The proposed model outperforms state-of-the-art models on a synthetic and real-world benchmark. |
Re-identification of De-identified Documents with Autoregressive Infilling (2025.acl-long)
Copied to clipboard
| Challenge: | Documents revealing sensitive information about individuals must typically be de-identified by masking all mentions of PII . |
| Approach: | They propose a method that attempts to reverse the process of re-identification based on a database of documents representing background knowledge. |
| Outcome: | The proposed method can recover as many as 80% of de-identified text spans and improve accuracy along with the level of background knowledge. |
Modeling Uncertainty in Composed Image Retrieval via Probabilistic Embeddings (2025.acl-long)
Copied to clipboard
Haomiao Tang, Jinpeng Wang, Yuang Peng, GuangHao Meng, Ruisheng Luo, Bin Chen, Long Chen, Yaowei Wang, Shu-Tao Xia
| Challenge: | Composed Image Retrieval (CIR) combines text and reference images to search for images . metric learning methods that focus on point embeddings fail to capture uncertainty in input data . |
| Approach: | They propose a framework that captures uncertainty in images and queries by Gaussian distributions in latent space rather than fixed points. |
| Outcome: | Experiments show that the proposed framework quantifies quality and semantic uncertainties . it can handle polysemy and ambiguity in search intentions, authors say . |
Untie the Knots: An Efficient Data Augmentation Strategy for Long-Context Pre-Training in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have prioritized expanding the context window from which they can incorporate more information. |
| Approach: | They propose a data augmentation strategy to enable large language models to gain long-context capabilities without the need to modify existing data mixture. |
| Outcome: | The proposed model outperforms existing models on 20 billion tokens and achieves 75% and 84.5% accuracy on RULER at 128K context length. |
APPL: A Prompt Programming Language for Harmonious Integration of Programs and Large Language Model Prompts (2025.acl-long)
Copied to clipboard
Honghua Dong, Qidong Su, Yubo Gao, Zhaoyu Li, Yangjun Ruan, Gennady Pekhimenko, Chris J. Maddison, Xujie Si
| Challenge: | Large Language Models (LLMs) are capable of handling diverse tasks with well-crafted prompts and integration of external tools. |
| Approach: | They propose a prompt programming language that acts as a bridge between computer programs and LLMs by providing convenient conventions between them. |
| Outcome: | The proposed language is intuitive, concise, and efficient through representative scenarios including Chain-of-Thought with self-consistency (CoT-SC) and ReAct tool-use agent. |
Evaluating Lexical Proficiency in Neural Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Natural Language Processing have been significantly shaped by the Deep Learning tsunami and the introduction of Transformer-based Language Models. |
| Approach: | They validated a framework to assess the lexical proficiency and linguistic creativity of Transformer-based Language Models (LMs) by analyzing performance of LMs of different sizes across tasks involving the generation, definition, and contextual usage of lexicals, neologisms, and nonce words. |
| Outcome: | The framework evaluates LMs in mono- and multilingual configuration across tasks involving the generation, definition, and contextual usage of lexicalized words, neologisms, and nonce words. |
Autoregressive Speech Synthesis without Vector Quantization (2025.acl-long)
Copied to clipboard
Lingwei Meng, Long Zhou, Shujie Liu, Sanyuan Chen, Bing Han, Shujie Hu, Yanqing Liu, Jinyu Li, Sheng Zhao, Xixin Wu, Helen M. Meng, Furu Wei
| Challenge: | MELLE is a novel language modeling approach for text-to-speech synthesis that generates continuous tokens from text . authors demonstrate that it reduces the need for vector quantization and improves model robustness . |
| Approach: | They propose to autoregressively generate continuous mel-spectrogram frames directly from text condition, bypassing vector quantization. |
| Outcome: | The proposed model achieves superior performance across multiple metrics and is more streamlined. |
Cuckoo: An IE Free Rider Hatched by Massive Nutrition in LLM’s Nest (2025.acl-long)
Copied to clipboard
| Challenge: | Massive high-quality data, both pre-training raw texts and post-training annotations, have been carefully prepared to incubate advanced large language models (LLMs). |
| Approach: | They propose to reframe next-token prediction into extraction for tokens already present in the context of LLMs by reframing next-tongue prediction into IE models. |
| Outcome: | The proposed model learns 102.6M extractive data converted from pre-training and post-training data with better performance than existing pre-trained IE models. |
FedEx-LoRA: Exact Aggregation for Federated and Efficient Fine-Tuning of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for low-rank averaging of LoRA adapters result in inexact updates. |
| Approach: | They propose a method which adds a residual error term to the pre-trained frozen weight matrix to achieve exact updates with minimal computational and communication overhead. |
| Outcome: | The proposed method achieves exact updates with minimal computational and communication overhead, preserving LoRA’s efficiency. |
Measuring Social Biases in Masked Language Models by Proxy of Prediction Quality (2025.acl-long)
Copied to clipboard
| Challenge: | Innovative transformer-based language models produce contextually-aware token embeddings, but have been shown to encode unwanted biases for downstream applications. |
| Approach: | They extend previous work by evaluating social biases introduced after retraining an MLM under the masked language modeling objective and propose proxy functions within an iterative masking experiment to measure the quality of transformer models’ predictions. |
| Outcome: | The proposed proxy functions within an iterative masking experiment show that all transformer models encode concerning social biases. |
Capturing Author Self Beliefs in Social Media Language (2025.acl-long)
Copied to clipboard
Siddharth Mangalik, Adithya V Ganesan, Abigail B. Wheeler, Nicholas Kerry, Jeremy D. W. Clifton, H. Schwartz, Ryan L. Boyd
| Challenge: | Existing methods for identifying self beliefs are limited. |
| Approach: | They propose a task that classifies language that contains explicit or implicit mentions of the author's self beliefs using an annotated set of 2,000 human-annotated self beliefs, 100,000 LLM-labeled examples, and 10,000 surveyed self belief paragraphs. |
| Outcome: | The proposed model outperforms OpenAI’s state-of-the-art GPT-4o model in the AUC of 0.944 and annotates 2,000 human-annotated self beliefs, 100,000 LLM-labeled examples, and 10,000 surveyed self belief paragraphs. |
Neural Topic Modeling with Large Language Models in the Loop (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated promising capabilities in topic discovery, but their direct application to topic modeling suffers from issues such as incomplete topic coverage, misalignment of topics, and inefficiency. |
| Approach: | They propose a novel LLM-in-the-loop framework that integrates Large Language Models with Neural Topic Models (NTMs) global topics and document representations are learned through the NTM, while an LLM refines these topics using an Optimal Transport (OT)-based alignment objective. |
| Outcome: | The proposed framework improves topic interpretability while preserving the efficiency of existing NTMs. |
HALoGEN: Fantastic LLM Hallucinations and Where to Find Them (2025.acl-long)
Copied to clipboard
| Challenge: | generative large language models produce hallucinations that are not aligned with world knowledge or input context. |
| Approach: | They propose a hallucination benchmark framework that measures hallucinism in large language models . they evaluate 150,000 generations from 14 language models and find they are riddled with hallucinos . |
| Outcome: | The proposed framework evaluates 150,000 generations from 14 language models. |
Synergizing LLMs with Global Label Propagation for Multimodal Fake News Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) can assist multimodal fake news detection by predicting pseudo labels, but their effective integration is non-trivial. |
| Approach: | They propose a global label propagation network with LLM-based pseudo labels for multimodal fake news detection which integrates LLM capabilities via label propagations. |
| Outcome: | The proposed model outperforms state-of-the-art models on benchmark datasets showing that it can propagate pseudo labels among all samples. |
“Yes, My LoRD.” Guiding Language Model Extraction with Locality Reinforced Distillation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for model extraction attacks on large language models are inadequate . existing methods neglect the inconsistency between training tasks and LLM alignment . |
| Approach: | They propose a model extraction algorithm that uses a policy-gradient-style training task to guide the crafting of preference for the local model. |
| Outcome: | The proposed algorithm reduces query complexity while mitigating watermark protection . it can extract various state-of-the-art commercial LLMs while minimizing query complexity . |
Jailbreak Large Vision-Language Models Through Multi-Modal Linkage (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to jailbreak large vision-language models fail against cutting-edge models such as GPT-4o, despite having undergone safety alignment training. |
| Approach: | They propose a new framework for jailbreaking large vision-language models that uses an encryption-decryption process to mitigate the over-exposure of harmful information. |
| Outcome: | The proposed framework jailbreaks GPT-4o with 99.40% success rates on SafeBench, 98.81% on MM-SafeBench and 99.07% on HADES-Dataset. |
Wait, that’s not an option: LLMs Robustness with Incorrect Multiple-Choice Options (2025.acl-long)
Copied to clipboard
| Challenge: | Using a framework that combines instruction-following with critical reasoning, we show that the ability of LLMs to override defaults when faced with invalid options is impaired by alignment techniques. |
| Approach: | They propose a framework for evaluating LLMs’ capacity to balance instruction-following with critical reasoning when presented with multiple-choice questions containing no valid answers. |
| Outcome: | The proposed framework improves models' ability to override defaults when faced with invalid options while minimizing the impact of model size and training techniques on the model. |
The Hidden Attention of Mamba Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have shown that Mamba models can be used for multiple domains, including NLP, long-range sequence processing, and computer vision. |
| Approach: | They add a third view and show that Mamba models can be viewed as attention-driven models. |
| Outcome: | The proposed model can be viewed as attention-driven and empirically compare it to the attention-based models of transformers. |
KV-Latent: Dimensional-level KV Cache Reduction with Frequency-aware Rotary Positional Embedding (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) based on Transformer Decoders have become the preferred choice for conversational generative AI. |
| Approach: | They propose a paradigm called KV-Latent to reduce the KV cache footprint and improve inference speed by down-sampling the Key-Value vector dimensions into a latent space. |
| Outcome: | The proposed paradigm reduces the KV Cache footprint and improves inference speed with a small amount of extra training, less than 1% of pre-training takes. |
LEANCODE: Understanding Models Better for Code Simplification of Pre-trained Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models often require significant computational resources, often constraining input word or code token lengths. |
| Approach: | They propose to use the encoder-decoder attention scores to represent the importance of a code token across multiple contexts to reduce training and prediction time. |
| Outcome: | The proposed approach outperforms the SOTAs DietCode and SlimCode in code search and summarization tasks. |
MARS: Benchmarking the Metaphysical Reasoning Abilities of Language Models with a Multi-task Evaluation Dataset (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in LLMs have demonstrated superior performance in a variety of reasoning tasks (Liu et al., 2023b; Chan e t al, 2024; Qin eetal., 2023) However, to truly achieve conscious processing, the integration of System II reasoning ability is essential. |
| Approach: | They propose a three-step process for reasoning with distributional changes, termed as a metaphysical resoning, and propose 'MARS' task to assess LLMs' reasoning abilities. |
| Outcome: | The proposed task is based on a three-step discriminative process and is compared with a standard model with 20 LLMs of varying sizes and methods. |
Ask-Before-Detection: Identifying and Mitigating Conformity Bias in LLM-Powered Error Detector for Math Word Problem Solutions (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have demonstrated the potential of large language models (LLMs) for automatic error detection in math word problems (MWPs). |
| Approach: | They propose a framework that generates adaptive reference solutions using LLMs to enhance error detection by reducing conformity bias in MWPs. |
| Outcome: | The proposed framework mitigates the performance gap between conventional and alternative solutions in MWPs, especially when combined with reasoning-enhancing techniques like chain-of-thought prompting. |
Real-time Factuality Assessment from Adversarial Feedback (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluations for assessing the factuality of news from conventional sources, such as claims on fact-checking websites, result in high accuracies over time for LLM-based detectors. |
| Approach: | They propose a pipeline that leverages natural language feedback from a RAG-based detector to iteratively modify real-time news into deceptive variants that challenge LLMs. |
| Outcome: | The proposed pipeline reduces the binary classification ROC-AUC by 17.5 percent for a strong RAG-based GPT-4o detector. |
Improve Vision Language Model Chain-of-thought Reasoning (2025.acl-long)
Copied to clipboard
Ruohong Zhang, Bowen Zhang, Yanghao Li, Haotian Zhang, Zhiqing Sun, Zhe Gan, Yinfei Yang, Ruoming Pang, Yiming Yang
| Challenge: | Current training recipes often rely on datasets dominated by short annotations with limited rationales, hindering the models' ability to generalize to tasks requiring comprehensive reasoning. |
| Approach: | They propose a two-stage post-training strategy that augments short answers with CoT reasoning generated by GPT-4o, enhancing the VLM's CoT capabilities through fine-tuning. |
| Outcome: | The proposed strategy enhances the model's CoT capabilities through fine-tuning and reinforcement learning. |
On the Mutual Influence of Gender and Occupation in LLM Representations (2025.acl-long)
Copied to clipboard
| Challenge: | We examine LLM representations of gender for first names in various occupational contexts to study how occupations and the gender perception of first names influence each other mutually. |
| Approach: | They examine LLM representations of gender for first names in various occupational contexts and examine how occupations and the gender perception of first names influence each other mutually. |
| Outcome: | The representations shift with the occupational context and are influenced by stereotypically feminine or masculine occupations. |
Disentangling Memory and Reasoning Ability in Large Language Models (2025.acl-long)
Copied to clipboard
Mingyu Jin, Weidi Luo, Sitao Cheng, Xinyi Wang, Wenyue Hua, Ruixiang Tang, William Yang Wang, Yongfeng Zhang
| Challenge: | Existing LLMs operate as an opaque process without explicit separation between knowledge retrieval and reasoning steps, making the decision-making process unclear and disorganized. |
| Approach: | They propose a language model inference paradigm that decomposes the complex inference process into two distinct and clear actions: (1) memory recall: which retrieves relevant knowledge, and (2) reasoning: which performs reasoning steps based on the recalled knowledge. |
| Outcome: | The proposed paradigm decomposes the inference process into two distinct and clear actions, memory and reason, guiding the model to distinguish between steps that require knowledge retrieval and those that involve reasoning. |
Open-World Attribute Mining for E-Commerce Products with Multimodal Self-Correction Instruction Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Current AM methods focus on extracting attributes from unimodal text, underutilizing multimodal data. |
| Approach: | They propose a framework for multimodal self-correction instruction tuning to extract new attributes from images and text with Multimodal Large Language Models. |
| Outcome: | The proposed framework outperforms state-of-the-art methods on two datasets. |
Normalized AOPC: Fixing Misleading Faithfulness Metrics for Feature Attributions Explainability (2025.acl-long)
Copied to clipboard
Joakim Edin, Andreas Geert Motzfeldt, Casper L. Christensen, Tuukka Ruotsalo, Lars Maaløe, Maria Maistro
| Challenge: | Deep neural network predictions are notoriously difficult to interpret due to the difficulty in understanding their inner mechanisms. |
| Approach: | They propose to normalize AOPC to enable consistent cross-model evaluations and more meaningful interpretation of individual scores. |
| Outcome: | The proposed approach can radically change AOPC results, questioning the conclusions of earlier studies and offering a more robust framework for assessing feature attribution faithfulness. |
Takin-VC: Expressive Zero-Shot Voice Conversion via Adaptive Hybrid Content Encoding and Enhanced Timbre Modeling (2025.acl-long)
Copied to clipboard
Yang Yuguang, Yu Pan, Jixun Yao, Xiang Zhang, Jianhao Ye, Hongbin Zhou, Lei Xie, Lei Ma, Jianjun Zhao
| Challenge: | Expressive zero-shot voice conversion (VC) aims to modify source timbre to match unseen speaker . existing zero- shot VC systems struggle to reproduce paralinguistic information in highly expressive speech . |
| Approach: | They propose a framework for expressive zero-shot voice conversion that uses hybrid content encoding and memory-augmented context-aware timbre modeling. |
| Outcome: | The proposed framework surpasses state-of-the-art VC systems in speech naturalness, speaker similarity, and speaker similarness. |
LangSAMP: Language-Script Aware Multilingual Pretraining (2025.acl-long)
Copied to clipboard
| Challenge: | Recent multilingual pretrained language models often avoid using language embeddings, which places a significant burden on token representations to encode all language-specific information. |
| Approach: | They propose a method that incorporates both language and script embeddings into the output of Transformer blocks before passing the final representations to the language modeling head for prediction. |
| Outcome: | The proposed method outperforms the baseline model in zero-shot crosslingual transfer across diverse downstream tasks. |
RelationalCoder: Rethinking Complex Tables via Programmatic Relational Transformation (2025.acl-long)
Copied to clipboard
| Challenge: | Semi-structured tables remain a major obstacle for automated data processing and analytics. |
| Approach: | They propose a technique called Loop Reference Decoding which identifies expandable groups and replicates each group using a concise loop over its repetitive region. |
| Outcome: | The proposed technique reduces output length from O(N M) to approximately O(K) Extensive experiments on HiTab and MultiHiertt show that it boosts Llama-2 and Mistral models by more than 20%, and GPT-4o by over 4%. |
Algorithmic Fidelity of Large Language Models in Generating Synthetic German Public Opinions: A Case Study (2025.acl-long)
Copied to clipboard
Bolei Ma, Berk Yoztyurk, Anna-Carolina Haensch, Xinpeng Wang, Markus Herklotz, Frauke Kreuter, Barbara Plank, Matthias Aßenmacher
| Challenge: | Recent advances in large language models have generated significant interest in their potential for synthetic data generation across various domains. |
| Approach: | They use open-ended survey data from the German Longitudinal Election Studies to prompt different LLMs to generate synthetic public opinions reflective of German subpopulations by incorporating demographic features into the persona prompts. |
| Outcome: | The LLM performs better for supporters of left-leaning parties like The Greens and The Left compared to other parties, and matches the least with the right-party AfD. |
TUNA: Comprehensive Fine-grained Temporal Understanding Evaluation on Dense Dynamic Videos (2025.acl-long)
Copied to clipboard
Fanheng Kong, Jingyuan Zhang, Hongzhi Zhang, Shi Feng, Daling Wang, Linhao Yu, Xingguang Ji, Yu Tian, V. W., Fuzheng Zhang
| Challenge: | Existing benchmarks for video understanding often focus on specific aspects, overlooking the holistic nature of video content. |
| Approach: | They propose a temporal-oriented benchmark for fine-grained understanding on dense dynamic videos with two complementary tasks: captioning and QA. |
| Outcome: | The proposed model performs well on diverse video scenarios and dynamic videos, with interpretable and robust evaluation criteria. |
Self-Instructed Derived Prompt Generation Meets In-Context Learning: Unlocking New Potential of Black-Box LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing prompt refinement methods suffer from semantic inconsistencies and fail to maintain users’ real intent. |
| Approach: | They propose a self-instructed in-context learning framework that generates reliable derived prompts while keeping semantic consistency with original prompts. |
| Outcome: | The proposed framework generates better derived prompts and significantly enhances LLMs’ ability to deliver more effective responses. |
Binary Classifier Optimization for Large Language Model Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for aligning large language models rely on preference-based approaches that require both positive and negative feedback as a pair. |
| Approach: | They propose a binary classifier optimization technique that trains a classifier using only binary feedback and a reward shift technique which minimizes the DPO loss. |
| Outcome: | The proposed method performs on a paired preference dataset and on 'likert-5 scale annotation dataset' it consistently demonstrates effective and robust alignment across four base LLMs and three different datasets, showcasing the strength of the proposed technique. |
UnSeenTimeQA: Time-Sensitive Question-Answering Beyond LLMs’ Memorization (2025.acl-long)
Copied to clipboard
Md Nayem Uddin, Amir Saeidi, Divij Handa, Agastya Seth, Tran Cao Son, Eduardo Blanco, Steven Corman, Chitta Baral
| Challenge: | UnSeenTimeQA is a data contamination-free time-sensitive question-answering benchmark. |
| Approach: | They propose a data contamination-free time-sensitive question-answering benchmark that avoids web-searchable queries grounded in the real world. |
| Outcome: | The proposed benchmark avoids web-searchable queries grounded in the real world and enables on-demand generation of new samples, mitigating the risk of data leakage. |
From Information to Insight: Leveraging LLMs for Open Aspect-Based Educational Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | a novel dataset summarizes student reflections on STEM lectures . ReflectASP eases the exploration of open-aspect-based summarization (OABS) despite the limitations of current datasets, it is still under-explored. |
| Approach: | They propose a dataset that summarizes student reflections on STEM lectures . they propose two refinement methods to improve summaries . |
| Outcome: | The proposed dataset summarizes student reflections on STEM lectures using automatic and human evaluations. |
AfriMed-QA: A Pan-African, Multi-Specialty, Medical Question-Answering Benchmark Dataset (2025.acl-long)
Copied to clipboard
Charles Nimo, Tobi Olatunji, Abraham Toluwase Owodunni, Tassallah Abdullahi, Emmanuel Ayodele, Mardhiyah Sanni, Ezinwanne C. Aka, Folafunmi Omofoye, Foutse Yuehgoh, Timothy Faniran, Bonaventure F. P. Dossou, Moshood O. Yekini, Jonas Kemp, Katherine A Heller, Jude Chidubem Omeke, Chidi Asuzu Md, Naome A Etori, Aïmérou Ndiaye, Ifeoma Okoh, Evans Doe Ocansey, Wendy Kinara, Michael L. Best, Irfan Essa, Stephen Edward Moore, Chris Fourie, Mercy Nyamewaa Asiedu
| Challenge: | Recent advances in large language models (LLMs) performance on medical multiplechoice question (MCQ) benchmarks have stimulated interest from healthcare providers and patients globally. |
| Approach: | They introduce AfriMed-QA, the first largescale Pan-African English multi-specialty medical Question-Answering (QA) dataset, with 15,000 questions sourced from over 60 medical schools across 16 countries. |
| Outcome: | The proposed model outperforms other models in the medical field and is compared with other models. |
Root Defense Strategies: Ensuring Safety of LLM at the Decoding Level (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to detect harmful outputs from prefill-level lacks utilization of the model’s decoding outputs, leading to relatively lower effectiveness and robustness. |
| Approach: | They propose a robust decoding mechanism that corrects harmful queries directly rather than rejecting them outright. |
| Outcome: | The proposed model improves model security without compromising reasoning speed. |
In-the-wild Audio Spatialization with Flexible Text-guided Localization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for mapping monaural audio to binaural signals lack flexibility and interactive control needed in complex multi-object user-interactive environments. |
| Approach: | They propose a text-guided audio spatialization framework that utilizes diverse text prompts to evaluate binaural audio models. |
| Outcome: | The proposed framework learns binaural differences guided by 3D spatial location and relative position prompts, enhanced with flipped-channel audio. |
L4Q: Parameter Efficient Quantization-Aware Fine-Tuning on Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Quantization-aware PEFT methods have been developed to reduce memory and computational costs associated with large language models. |
| Approach: | They propose a method that integrates Quantization-Aware Training (QAT) with LoRA to reduce memory overhead and improve model accuracy. |
| Outcome: | The proposed method significantly reduces QAT’s memory overhead while preserving the advantage of QAT in producing fully quantized LLMs with high accuracy. |
Second Language (Arabic) Acquisition of LLMs via Progressive Vocabulary Expansion (2025.acl-long)
Copied to clipboard
Jianqing Zhu, Huang Huang, Zhihang Lin, Juhao Liang, Zhengyang Tang, Khalid Almubarak, Mosen Alharthi, Bang An, Juncai He, Xiangbo Wu, Fei Yu, Junying Chen, Ma Zhuoheng, Yuhao Du, He Zhang, Saied Alshahrani, Emad A. Alghamdi, Lian Zhang, Ruoyu Sun, Haizhou Li, Benyou Wang, Jinchao Xu
| Challenge: | In the evolving landscape of large language models, the predominant focus has been on English and Chinese. |
| Approach: | They propose to utilize Arabic-specific vocabulary in the tokenizer to accelerate decoding. |
| Outcome: | The proposed model achieves decent performance comparable to the best Arabic LLMs across various Arabic benchmarks. |
What Really Matters in Many-Shot Attacks? An Empirical Study of Long-Context Vulnerabilities in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in context length extension have improved LLMs' performance and safety, but they present critical safety challenges. |
| Approach: | They investigate long-context vulnerabilities in Large Language Models (LLMs) using many-shot jailbreaking to exploit context length extension. |
| Outcome: | The proposed attacks do not require carefully crafted harmful content. |
ECERC: Evidence-Cause Attention Network for Multi-Modal Emotion Recognition in Conversation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for multi-modal emotion recognition in isolated utterances do not capture emotional causes, including emotional contagion, influences from others, and self-referenced or externally introduced events. |
| Approach: | They propose a multi-modal conversational emotion recognition system that integrates emotional evidence with contextual causes through five stages. |
| Outcome: | The proposed method achieves competitive performance on two widely used benchmark datasets, IEMOCAP and MELD. |
CompileAgent: Automated Real-World Repo-Level Compilation with Tool-Integrated LLM-based Agent System (2025.acl-long)
Copied to clipboard
Li Hu, Guoqiang Chen, Xiuwei Shang, Shaoyin Cheng, Benlong Wu, LiGangyang LiGangyang, Xu Zhu, Weiming Zhang, Nenghai Yu
| Challenge: | CompileAgent is the first LLM-based agent framework dedicated to repo-level compilation. |
| Approach: | They propose a LLM-based agent framework dedicated to repo-level compilation. |
| Outcome: | The proposed method significantly improves compilation success rate, ranging from 10% to 71%. |
Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals’ Subjective Text Perceptions (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work has shown that LLMs perform poorly when prompted with sociodemographic attributes, suggesting limited inherent sociodemography knowledge. |
| Approach: | They propose to train large language models to be accurate sociodemographic models of annotator variation by using a curated dataset of five tasks with standardized sociodemography. |
| Outcome: | The proposed models improve in sociodemographic prompting when trained but this performance gain is largely due to models learning annotator-specific behaviour rather than sociodemography. |
Exploring Forgetting in Large Language Model Pre-Training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research on task-level forgetting in LLMs has focused on pretraining . but, there is limited attention to finer-grained forgetting during training . |
| Approach: | They investigated the existence and measurement of forgetting in pre-training . they examined low-cost, straightforward methods to mitigate forgetting during the pre- training phase . |
| Outcome: | The proposed methods could be used to mitigate forgetting during the pre-training phase and offer insights into the dynamics of forgetting. |
Bias in the Mirror : Are LLMs opinions robust to their own adversarial attacks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on large language models lacks robustness, highlighting the limitations of such models. |
| Approach: | They propose a novel approach where two LLMs engage in self-debate to persuade a neutral version of the model. |
| Outcome: | The proposed approach examines whether large language models are robust during interactions and whether they are susceptible to reinforcing misinformation or shifting to harmful viewpoints. |
AndroidLab: Training and Systematic Benchmarking of Android Autonomous Agents (2025.acl-long)
Copied to clipboard
Yifan Xu, Xiao Liu, Xueqiao Sun, Siyi Cheng, Hao Yu, Hanyu Lai, Shudan Zhang, Dan Zhang, Jie Tang, Yuxiao Dong
| Challenge: | Existing studies on Android agents lack systematic research on open-source and closed-source models. |
| Approach: | They propose a framework for Android agents that includes an operation environment and a reproducible benchmark. |
| Outcome: | The proposed framework lifts the success rate of open-source LLMs and LMMs from 4.59% to 21.50% for LLM and 1.93% to 13.28% for LMM. |
Modular Sentence Encoders: Separating Language Specialization from Cross-Lingual Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Multilingual sentence encoders are often trained to map sentences from different languages into a shared semantic vector space . cross-lingual alignment training distorts optimal monolingual structure of semantic spaces of individual languages . a modular solution can be used for cross-linguistic tasks such as cross-language semantic similarity and zero-shot transfer . |
| Approach: | They propose a modular training system that embeds sentences from different languages into a shared semantic vector space. |
| Outcome: | The proposed solution achieves better performance across all tasks compared to monolithic models. |
Multimodal Transformers are Hierarchical Modal-wise Heterogeneous Graphs (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal Sentiment Analysis (MSA) is a rapidly developing field that integrates multimodal information to recognize sentiments. |
| Approach: | They propose a multimodal fusion model that integrates multimodal information to recognize sentiments using multimodal transformers. |
| Outcome: | The proposed model achieves significantly higher performance than MulTs and the existing model is robust. |
Have We Designed Generalizable Structural Knowledge Promptings? Systematic Evaluation and Rethinking (2025.acl-long)
Copied to clipboard
Yichi Zhang, Zhuo Chen, Lingbing Guo, Yajing Xu, Shaokai Chen, Mengshu Sun, Binbin Hu, Zhiqiang Zhang, Lei Liang, Wen Zhang, Huajun Chen
| Challenge: | Existing methods to integrate external knowledge into LLMs focus on specific problems, lacking a comprehensive exploration of the generalization and capability boundaries of SKP. |
| Approach: | They propose a new paradigm for structural knowledge prompting to integrate external structural knowledge into LLMs by incorporating structural representations. |
| Outcome: | The proposed benchmark SUBARU enables the evaluation of the generalization capabilities of SKP from four perspectives. |
LLäMmlein: Transparent, Compact and Competitive German-Only Language Models from Scratch (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have achieved remarkable success, yet this progress is predominantly centered on English. |
| Approach: | They create two German-only decoder models from scratch and publish them for the (German) NLP research community to use. |
| Outcome: | The two models performed competitively on the German SuperGLEBer benchmark, but performance improvements plateaued early during training, offering valuable insights into resource allocation for future models. |
Speaking Beyond Language: A Large-Scale Multimodal Dataset for Learning Nonverbal Cues from Video-Grounded Dialogues (2025.acl-long)
Copied to clipboard
Youngmin Kim, Jiwan Chung, Jisoo Kim, Sunghyun Lee, Sangkyu Lee, Junhyeok Kim, Cheoljong Yang, Youngjae Yu
| Challenge: | Existing large language models fail to incorporate nonverbal elements into conversational experiences. |
| Approach: | They propose a multimodal language model that generates nonverbal cues alongside text . their dataset is annotated with time-aligned text, facial expressions, and body language . |
| Outcome: | The proposed model generates nonverbal languages and text, corresponding to conversational input. |
How Much Do Encoder Models Know About Word Senses? (2025.acl-long)
Copied to clipboard
| Challenge: | Word Sense Disambiguation (WSD) is a key task in Natural Language Processing (NLP) however, how well these models inherently disambiguate word senses remains uncertain. |
| Approach: | They evaluate several encoder-only PLMs across WordNet and ODE sense inventories to evaluate their ability to separate word senses without any task-specific fine-tuning. |
| Outcome: | The proposed model outperforms output layer on WordNet and ODE sense inventories by 15 percentage points. |
When Backdoors Speak: Understanding LLM Backdoor Attacks Through Model-Generated Explanations (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have shown that Large Language Models (LLMs) are susceptible to backdoor attacks, where triggers embedded in poisoned data can maliciously alter LLMs’ behaviors. |
| Approach: | They propose to leverage LLMs' generative capabilities to generate human-readable explanations for their decisions, enabling direct comparisons between explanations of clean and poisoned data. |
| Outcome: | The proposed model produces coherent explanations for clean inputs but logically flawed explanations on poisoned data. |
HateDay: Insights from a Global Hate Speech Dataset Representative of a Day on Twitter (2025.acl-long)
Copied to clipboard
Manuel Tonneau, Diyi Liu, Niyati Malhotra, Scott A. Hale, Samuel Fraiberger, Victor Orozco-Olvera, Paul Röttger
| Challenge: | Prior work on automated hate speech detection models has been limited due to systematic biases in evaluation datasets and poor performance across geographies. |
| Approach: | They propose to construct a global hate speech dataset representative of social media settings from tweets posted on September 21, 2022. |
| Outcome: | The proposed dataset covers eight languages and four English-speaking countries and covers eight countries where English is the main language on Twitter. |
LegalAgentBench: Evaluating LLM Agents in Legal Domain (2025.acl-long)
Copied to clipboard
Haitao Li, Junjie Chen, Jingli Yang, Qingyao Ai, Wei Jia, Youfeng Liu, Kai Lin, Yueyue Wu, Guozhi Yuan, Yiran Hu, Wuyue Wang, Yiqun Liu, Minlie Huang
| Challenge: | Existing general-domain benchmarks do not capture complexity of real-world judicial cognition and decision-making. |
| Approach: | They propose a benchmark specifically designed to evaluate LLM Agents in the legal domain. |
| Outcome: | The proposed benchmark includes 17 corpora from real-world legal scenarios and provides 37 tools for interacting with external knowledge. |
Inference Compute-Optimal Video Vision Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Using video vision language models, inference costs are often more expensive than finetuning. |
| Approach: | They investigate the optimal allocation of inference compute across three key scaling factors in video vision language models. |
| Outcome: | The proposed model configurations are based on three key scaling factors . the results can be applied to real-world tasks and tasks with fixed inference budgets. |
Steering into New Embedding Spaces: Analyzing Cross-Lingual Alignment Induced by Model Interventions in Multilingual Language Models (2025.acl-long)
Copied to clipboard
Anirudh Sundar, Sinead Williamson, Katherine Metcalf, Barry-John Theobald, Skyler Seto, Masha Fedzechkina
| Challenge: | Large language models (LLMs) exhibit impressive performance on a variety of tasks from text summarization to zero-shot common-sense reasoning. |
| Approach: | They propose to manipulate the embedding space of mLLMs by manipulating its activations to steer generation into the desired direction. |
| Outcome: | The proposed model interventions improves alignment of cross-lingual representations in multilingual large language models with up to 2x improvements in top-1 accuracy on cross-linguistic retrieval tasks. |
Digital Gatekeepers: Google’s Role in Curating Hashtags and Subreddits (2025.acl-long)
Copied to clipboard
| Challenge: | This study examines how search engines like Google selectively promote or suppress certain hashtags and subreddits, impacting the flow of information and impacting public conversations. |
| Approach: | They compare search engine results with nonsampled data from Reddit and Twitter/X to examine how search engines curate content through algorithmic curation. |
| Outcome: | The proposed algorithm suppresses subreddits related to sexually explicit material, conspiracy theories, advertisements, and cryptocurrencies while promoting content associated with higher engagement. |
Behind Closed Words: Creating and Investigating the forePLay Annotated Dataset for Polish Erotic Discourse (2025.acl-long)
Copied to clipboard
| Challenge: | specialized Polish language models are more effective at detecting harmful content than traditional methods. |
| Approach: | They propose a Polish-language dataset for erotic content detection that captures ambiguity, violence, and socially unacceptable behaviors. |
| Outcome: | The proposed dataset shows that specialized Polish language models achieve superior performance compared to multilingual alternatives, with transformer-based architectures showing particular strength in handling imbalanced categories. |
Assessment and manipulation of latent constructs in pre-trained language models using psychometric scales (2025.acl-long)
Copied to clipboard
Maor Reuben, Ortal Slobodin, Idan-Chaim Cohen, Aviad Elyashar, Orna Braun-Lewensohn, Odeya Cohen, Rami Puzis
| Challenge: | a recent study suggests that language models may be tricked into answering psychometric questionnaires, but they cannot be assessed because of inadequate psychometric methods. |
| Approach: | They propose to re-form standard psychological questionnaires into natural language inference prompts and a code library to support the psychometric assessment of arbitrary models. |
| Outcome: | The proposed model can be reformulated into natural language inference prompts and a code library to support the psychometric assessment of arbitrary models. |
Did Translation Models Get More Robust Without Anyone Even Noticing? (2025.acl-long)
Copied to clipboard
| Challenge: | Neural machine translation models are highly sensitive to “noisy” inputs, such as spelling errors, abbreviations, and formatting issues. |
| Approach: | They revisit this insight in light of recent multilingual MT models and large language models applied to machine translation. |
| Outcome: | The proposed models perform better on clean data than previous models, but none of the open models use robustness techniques. |
Nemotron-CC: Transforming Common Crawl into a Refined Long-Horizon Pretraining Dataset (2025.acl-long)
Copied to clipboard
Dan Su, Kezhi Kong, Ying Lin, Joseph Jennings, Brandon Norick, Markus Kliegl, Mostofa Patwary, Mohammad Shoeybi, Bryan Catanzaro
| Challenge: | Recent Common Crawl datasets remove 90% of data, limiting their suitability for long token horizon training. |
| Approach: | They propose to combine classifier ensembling, synthetic data rephrasing and heuristic filters to achieve better trade-offs between accuracy and data quantity. |
| Outcome: | The proposed model-based filtering improves MMLU by 5.6 over DCLM for 15T tokens . the full 6.3T token dataset matches DCLM on MMLO, but contains four times more unique real tokens than DCLM . |
Hierarchical Level-Wise News Article Clustering via Multilingual Matryoshka Embeddings (2025.acl-long)
Copied to clipboard
| Challenge: | Contextual large language model embeddings are often monolingual, do not scale, and struggle in multilingual settings. |
| Approach: | They propose a hierarchical approach to embed news articles and social media data using Matryoshka embeddings that can determine story similarity at varying levels of granularity based on which subset of dimensions is examined. |
| Outcome: | The proposed model achieves state-of-the-art performance on the SemEval 2022 task 8 dataset. |
Contrastive Perplexity for Controlled Generation: An Application in Detoxifying Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to generate toxic content by large language models are based on pipelines . current approaches focus on preserving performance while effectively mitigating toxicity . |
| Approach: | They propose a framework for implicit knowledge editing and controlled text generation by using hard negatives. |
| Outcome: | The proposed framework significantly reduces toxic generation while maintaining strong performance on downstream tasks. |
INVESTORBENCH: A Benchmark for Financial Decision-Making Tasks with LLM-based Agent (2025.acl-long)
Copied to clipboard
Haohang Li, Yupeng Cao, Yangyang Yu, Shashidhar Reddy Javaji, Zhiyang Deng, Yueru He, Yuechen Jiang, Zining Zhu, K.p. Subbalakshmi, Jimin Huang, Lingfei Qian, Xueqing Peng, Jordan W. Suchow, Qianqian Xie
| Challenge: | Recent advances have underscored the potential of large language model (LLM)-based agents in financial decision-making. |
| Approach: | They propose to evaluate LLM agents using 13 different LLMs as backbone models across various market environments and tasks. |
| Outcome: | The proposed framework assesses the reasoning and decision-making capabilities of 13 different LLMs across various market environments and tasks. |
Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference (2025.acl-long)
Copied to clipboard
Benjamin Warner, Antoine Chaffin, Benjamin Clavié, Orion Weller, Oskar Hallström, Said Taghadouini, Alexis Gallagher, Raja Biswas, Faisal Ladhak, Tom Aarsen, Griffin Thomas Adams, Jeremy Howard, Iacopo Poli
| Challenge: | Encoder-only transformer models such as BERT offer a great performance-size tradeoff for retrieval and classification tasks compared to larger decoder models. |
| Approach: | They introduce a new transformer model, ModernBERT, which brings modern model optimizations to encoder-only transformer models. |
| Outcome: | The proposed model improves on the BERT transformer model and is faster and more memory efficient than the older models. |
Gender Inclusivity Fairness Index (GIFI): A Multilevel Framework for Evaluating Gender Diversity in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | GIFI measures the diversity of LLMs' outputs, including gender identifiers, and identifies gender biases associated with varying gender identifiers. |
| Approach: | They propose a Gender Inclusivity Fairness Index (GIFI) that quantifies the diverse gender inclusivity of large language models. |
| Outcome: | The proposed metric quantifies the diversity of LLMs across multiple dimensions, including non-binary identities. |
D.Va: Validate Your Demonstration First Before You Use It (2025.acl-long)
Copied to clipboard
| Challenge: | In-context learning (ICL) heavily relies on selecting effective demonstrations to achieve outputs that better align with the expected results. |
| Approach: | They propose a method which integrates a demonstration validation perspective into this field and integrates it into the learning paradigm. |
| Outcome: | The proposed method surpasses all retrieval-based in-context learning techniques across both natural language understanding (NLU) and natural language generation (NLG) tasks. |
Are Any-to-Any Models More Consistent Across Modality Transfers Than Specialists? (2025.acl-long)
Copied to clipboard
| Challenge: | equivariance evaluations uncover weak but observable consistency through structured analyses of the intermediate latent space enabled by multiple editing operations. |
| Approach: | They use a dataset of 1,000 images paired with captions, editing instructions, and Q&A pairs to evaluate cross-modal transfers rigorously. |
| Outcome: | The proposed models do not consistently demonstrate greater cross-modal consistency than specialized models in pointwise evaluations such as cyclic consistency. |
MAIN-RAG: Multi-Agent Filtering Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
Chia-Yuan Chang, Zhimeng Jiang, Vineeth Rakesh, Menghai Pan, Chin-Chia Michael Yeh, Guanchu Wang, Mingzhi Hu, Zhichao Xu, Yan Zheng, Mahashweta Das, Na Zou
| Challenge: | Existing RAG systems struggle with the quality of retrieval documents, causing performance degradation and reducing performance. |
| Approach: | They propose a training-free RAG framework that leverages multiple LLM agents to collaboratively filter and score retrieved documents. |
| Outcome: | The proposed framework outperforms existing RAG frameworks in QA benchmarks and shows superior answer consistency and answer accuracy over baseline methods. |
Unraveling the Mechanics of Learning-Based Demonstration Selection for In-Context Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Recent learning-based demonstration selection methods have proven beneficial to in-context learning (ICL) by choosing more useful exemplars. |
| Approach: | They propose two methods to capture task-agnostic similarities between input and output of LLMs. |
| Outcome: | The proposed methods integrate task-agnostic similarities of different levels between input and output of exemplars and test cases to eliminate costly data collection. |
Direct Prompt Optimization with Continuous Representations (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for prompt optimization for language models lack extensibility and search space. |
| Approach: | They propose a method that integrates greedy strategies into optimization with continuous representations to address instability caused by rounding. |
| Outcome: | The proposed approach can improve prompt optimization performance on text classification and attack tasks, as well as models, including GPT-2, OPT, Vicuna, and LLaMA-2. |
uMedSum: A Unified Framework for Clinical Abstractive Summarization (2025.acl-long)
Copied to clipboard
Aishik Nagar, Yutong Liu, Andy T. Liu, Viktor Schlegel, Vijay Prakash Dwivedi, Arun-Kumar Kaliya-Perumal, Guna Pratheep Kalanchiam, Yili Tang, Robby T. Tan
| Challenge: | Clinical abstractive summarization struggles to balance faithfulness and informativeness, sacrificing key information or introducing confabulations. |
| Approach: | They develop a modular hybrid framework that integrates confabulation removal and key information addition into abstractive summarization methods. |
| Outcome: | The proposed framework outperforms state-of-the-art abstractive summarization methods in both quantitative metrics and expert evaluations. |
GigaSpeech 2: An Evolving, Large-Scale and Multi-domain ASR Corpus for Low-Resource Languages with Automated Crawling, Transcription and Refinement (2025.acl-long)
Copied to clipboard
Yifan Yang, Zheshu Song, Jianheng Zhuo, Mingyu Cui, Jinpeng Li, Bo Yang, Yexing Du, Ziyang Ma, Xunying Liu, Ziyuan Wang, Ke Li, Shuai Fan, Kai Yu, Wei-Qiang Zhang, Guoguo Chen, Xie Chen
| Challenge: | GigaSpeech 2 is a large-scale, multi-domain, multilingual speech recognition corpus for low-resource languages. |
| Approach: | They propose a large-scale, multi-domain, multilingual speech recognition corpus for low-resource languages and an automated pipeline for data crawling, transcription, and label refinement. |
| Outcome: | The proposed corpus reduces the word error rate for Thai, Indonesian, and Vietnamese on a realistic YouTube test set by 25% to 40% compared to Whisper large-v3. |
Context-Aware Sentiment Forecasting via LLM-based Multi-Perspective Role-Playing Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to predict sentiments on social media are limited and do not consider reciprocal influences among social media users. |
| Approach: | They propose a multi-perspective role-playing framework to simulate human response processes to extract sentiment-related features from social media messages. |
| Outcome: | The proposed model improves sentiment forecasting at microscopic and macroscopic levels. |
TARGA: Targeted Synthetic Data Generation for Practical Reasoning over Structured Data (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for semantic parsing rely on extensive manually annotated datasets and limited generalization capability to unseen examples. |
| Approach: | They propose a framework that generates high-relevance synthetic data without manual annotation . they generate queries for the queries and use them as demonstrations for in-context learning . |
| Outcome: | The proposed framework outperforms non-fine-tuned methods on KBQA datasets and shows superior sample efficiency, robustness, and generalization capabilities under non-I.I.D. settings. |
AndroidGen: Building an Android Language Agent under Data Scarcity (2025.acl-long)
Copied to clipboard
| Challenge: | Existing LLMs lack high-quality data sources and lack robust data filtration strategies. |
| Approach: | They develop a framework to enhance the capabilities of LLM-based agents under data scarcity. |
| Outcome: | The proposed framework improves the capabilities of LLM-based agents under data scarcity. |
Prompt Candidates, then Distill: A Teacher-Student Framework for LLM-driven Data Annotation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for data annotation use an aggressive approach prompting LLMs to determine a single gold label for each unlabeled sample. |
| Approach: | They propose a teacher-student framework that distills candidate annotations with a Small Language Model (SLM) they propose to use LLMs to generate and distill candidate annotation with slms to ensure unique labels are provided for downstream tasks. |
| Outcome: | The proposed method outperforms existing methods due to uncertainty in LLMs and is noisetolerant. |
A Survey of Post-Training Scaling in Large Language Models (2025.acl-long)
Copied to clipboard
Hanyu Lai, Xiao Liu, Junjie Gao, Jiale Cheng, Zehan Qi, Yifan Xu, Shuntian Yao, Dan Zhang, Jinhua Du, Zhenyu Hou, Xin Lv, Minlie Huang, Yuxiao Dong, Jie Tang
| Challenge: | Large language models (LLMs) have demonstrated proficiency in understanding and generating human natural languages. |
| Approach: | They propose a framework for scaling large language models using supervised fine-tuning, RLxF and test-time compute methodologies. |
| Outcome: | The proposed model can be used to understand and generate human natural languages. |
Position-aware Automatic Circuit Discovery (2025.acl-long)
Copied to clipboard
| Challenge: | Existing circuit analysis methods treat component positions as equally relevant across input positions, limiting their ability to capture cross-positional interactions or mechanisms that vary across positions. |
| Approach: | They propose to extend edge attribution patching to differentiate between token positions and introduce a dataset schema to enable position-aware circuit discovery in datasets with variable length examples. |
| Outcome: | The proposed approach enables position-aware circuit discovery even on tasks with variable length examples. |
HyperFM: Fact-Centric Multimodal Fusion for Link Prediction over Hyper-Relational Knowledge Graphs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing link prediction techniques focus on learning the complex relationships between entities and relations while ignoring the multimodal information. |
| Approach: | They propose a fact-centric fusion technique that captures complex interactions between different data modalities while accommodating the hyper-relational structure of the KG in a facts-centric manner. |
| Outcome: | The proposed technique improves on two real-world KG datasets by 6.0-6.8% over baselines. |
Centurio: On Drivers of Multilingual Ability of Large Vision-Language Model (2025.acl-long)
Copied to clipboard
Gregor Geigle, Florian Schneider, Carolin Holtermann, Chris Biemann, Radu Timofte, Anne Lauscher, Goran Glavaš
| Challenge: | Existing models for large vision-language tasks are trained on English data, which makes them struggle to understand non-English input and fail to generate output in the desired target language. |
| Approach: | They conduct multi-stage experiments on 13 vision-language tasks and 43 languages . they find that one can include as many as 100 training languages simultaneously with as little as 25-50% of non-English data . |
| Outcome: | The proposed model outperforms existing models in 14 tasks and 56 languages. |
Less for More: Enhanced Feedback-aligned Mixed LLMs for Molecule Caption Generation and Fine-Grained NLI Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent trends have led to the use of multimodal models to learn molecular and linguistic representations, either in separate but coordinated spaces or in a common space. |
| Approach: | They propose a novel atomic-level evaluation method leveraging off-the-shelf Natural Language Inference (NLI) models for use in the unseen chemical domain. |
| Outcome: | The proposed method surpasses state-of-the-art models in the unseen chemical domain while relying on a granularity-based evaluation method. |
Ensemble Watermarks for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing watermarks for large language models lack flexibility and struggle with attacks such as paraphrasing. |
| Approach: | They propose a multi-feature method for generating watermarks that combines multiple distinct watermark features into an ensemble watermark. |
| Outcome: | The proposed method achieves 98% detection rate after paraphrasing attack and 49% after paraphrase attack. |
\mathsf{Con Instruction}: Universal Jailbreaking of Multimodal Large Language Models via Non-Textual Modalities (2025.acl-long)
Copied to clipboard
| Challenge: | Existing attacks communicate instruction through text, accompanied by a toxic image or audio . a novel gray-box attack method generates adversarial images or audio to convey harmful instructions to MLLMs . |
| Approach: | They propose a gray-box attack method that generates adversarial images or audio to convey specific harmful instructions to MLLMs by following non-textual instruction. |
| Outcome: | The proposed method achieves highest success rates on visual and audio-language models . larger models are more susceptible toCon Instruction, compared to their underlying models - the results will be released . |
TRACT: Regression-Aware Fine-tuning Meets Chain-of-Thought Reasoning for LLM-as-a-Judge (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning LLMs use cross-entropy (CE) loss . existing methods neglect the numeric nature of score prediction . |
| Approach: | They propose a method that fine-tunes large language models (LLMs) for automated text evaluation, assigning a score to the input based on scoring rubrics. |
| Outcome: | The proposed model outperforms existing methods in four LLM-as-a-judge datasets and two LLMs. |
DioR: Adaptive Cognitive Detection and Contextual Retrieval Optimization for Dynamic Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for generating large language models face limitations in key aspects such as retrieval triggers and contextual scrutiny of retrieval content. |
| Approach: | They propose a dynamic RAG method that uses cognitive detection and contextual retrieval optimization to determine when retrieval is needed and what to retrieve for LLMs. |
| Outcome: | The proposed method achieves superior performance on all tasks, demonstrating the effectiveness of the proposed method. |
Unveiling the Power of Source: Source-based Minimum Bayes Risk Decoding for Neural Machine Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Maximum a posteriori decoding aims to maximize the estimated posterior probability, but high estimated probability does not always lead to high translation quality. |
| Approach: | They propose a method that seeks hypotheses with the highest expected utility by using quasi-sources as “support hypothese . they propose sMBR decoding which utilizes a reference-free quality estimation metric as the utility function. |
| Outcome: | The proposed approach outperforms QE reranking and the standard MBR decoding. |
ToolHop: A Query-Driven Benchmark for Evaluating Large Language Models in Multi-Hop Tool Use (2025.acl-long)
Copied to clipboard
Junjie Ye, Zhengyin Du, Xuesong Yao, Weijian Lin, Yufei Xu, Zehui Chen, Zaiyuan Wang, Sining Zhu, Zhiheng Xi, Siyu Yuan, Tao Gui, Qi Zhang, Xuanjing Huang, Jiecao Chen
| Challenge: | Effective evaluation of multi-hop tool use is critical for analyzing the understanding, reasoning, and function-calling capabilities of large language models. |
| Approach: | They propose a dataset that provides rigorous evaluation of multi-hop tool use. |
| Outcome: | The proposed model achieves 49.04% accuracy across five model families. |
Mixture of insighTful Experts (MoTE): The Synergy of Reasoning Chains and Expert Mixtures in Self-Alignment (2025.acl-long)
Copied to clipboard
Zhili Liu, Yunhao Gou, Kai Chen, Lanqing Hong, Jiahui Gao, Fei Mi, Yu Zhang, Zhenguo Li, Xin Jiang, Qun Liu, James Kwok
| Challenge: | Recent studies show that reasoning abilities contribute significantly to model safety, while integrating Mixture-of-Experts (MoE) architectures can further enhance alignment. |
| Approach: | They propose a framework that synergistically combines reasoning chains and expert mixtures to improve self-alignment. |
| Outcome: | The proposed framework improves model safety, jailbreak resistance, and over-refusal capabilities, achieving performance comparable to OpenAI’s state-of-the-art o1 model. |
MAPS: Motivation-Aware Personalized Search via LLM-Driven Consultation Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing personalized product search methods assume that users’ query fully captures their real motivation, but in practice, user's queries do not always articulate the requirements. |
| Approach: | They propose a Motivation-Aware Personalized Search method that embeds queries and consultations into a unified semantic space via LLMs and utilizes a Mixture of Attention Experts (MoAE) to prioritize critical semantics. |
| Outcome: | Extensive experiments on real and synthetic data show that the proposed method outperforms existing methods in retrieval and ranking tasks. |
Aristotle: Mastering Logical Reasoning with A Logic-Complete Decompose-Search-Resolve Framework (2025.acl-long)
Copied to clipboard
Jundong Xu, Hao Fei, Meng Luo, Qian Liu, Liangming Pan, William Yang Wang, Preslav Nakov, Mong-Li Lee, Wynne Hsu
| Challenge: | Existing systems fail to fully leverage the structure of logical tasks throughout the reasoning process, causing bottlenecks in efficiency and efficacy. |
| Approach: | They propose a logic-complete reasoning framework, Aristotle, which integrates symbolic expressions and logical rules into the entire reasoning process. |
| Outcome: | The proposed framework outperforms state-of-the-art reasoning frameworks in accuracy and efficiency. |
LADM: Long-context Training Data Selection with Attention-based Dependency Measurement for LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Long-context modeling has drawn more attention in the area of Large Language Models (LLMs). |
| Approach: | They propose a Long-context data selection framework with Attention-based Dependency Measurement which can efficiently identify high-quality long-contrast data from a large-scale, multi-domain pre-training corpus. |
| Outcome: | The proposed framework significantly boosts the performance of LLMs on multiple long-context tasks with only 1B tokens for continual training. |
Iron Sharpens Iron: Defending Against Attacks in Machine-Generated Text Detection with Adversarial Training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing MGT detectors are vulnerable to simple perturbations and adversarial attacks. |
| Approach: | They propose an adversarial framework for training a robust machine-generated text detector called GREedy Adversary PromoTed DefendER. |
| Outcome: | The proposed framework reduces the Attack Success Rate (ASR) by 0.67% compared with SOTA defense methods. |
Cultural Learning-Based Culture Adaptation of Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches for adapting large language models to diverse cultural values often rely on prompt engineering. |
| Approach: | They propose a framework for enhancing LLM alignment with cultural values based on cultural learning that leverages simulated social interactions to generate role-playing scenarios. |
| Outcome: | The proposed framework improves cultural value alignment across various model architectures measured using World Value Survey data. |
A-TASC: Asian TED-Based Automatic Subtitling Corpus (2025.acl-long)
Copied to clipboard
| Challenge: | Existing AS corpora and primary metric SubER focus on European languages. |
| Approach: | They propose an Asian TED-based automatic subtitling corpus derived from English TED Talks and a modification of SubER to enable reliable evaluation of subtitle quality for languages without explicit word boundaries. |
| Outcome: | The proposed corpus is based on TED Talks audio segments, transcripts, and subtitles in Chinese, Japanese, Korean, and Vietnamese. |
Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal Training (2025.acl-long)
Copied to clipboard
Youliang Yuan, Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Jiahao Xu, Tian Liang, Pinjia He, Zhaopeng Tu
| Challenge: | Large Language Models exhibit a level of intelligence that is both impressive and everevolving, but their ability to refuse generating unsafe content is a double-edged sword. |
| Approach: | They propose a method to tackle a refusal position bias within safety tuning data that compromises the models’ ability to appropriately refuse generating unsafe content. |
| Outcome: | The proposed method significantly improves model safety without compromising performance and surpasses baseline methods in defending against attacks. |
Token Prepending: A Training-Free Approach for Eliciting Better Sentence Embeddings from LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have focused on prompt engineering to extract sentence embeddings from large language models (LLMs) but these models are mostly decoder-only and the earlier tokens in the sentence cannot attend to the latter, resulting in biased encoding of sentence information and cascading effects on the final decoded token. |
| Approach: | They propose a plug-and-play and training-free technique that prepends each layer’s decoded sentence embedding to the beginning of the sentence in the next layer’ s input. |
| Outcome: | The proposed technique can significantly improve the performance of existing prompt-based sentence embedding methods across different LLMs while incurring negligible additional inference cost. |
No Questions are Stupid, but some are Poorly Posed: Understanding Poorly-Posed Information-Seeking Questions (2025.acl-long)
Copied to clipboard
| Challenge: | When a question is poorly posed, answerers struggle to converge on dominant interpretations, while models attempt comprehensive coverage by addressing many interpretations simultaneously. |
| Approach: | They propose a computational framework to study poorly-posedness of questions by generating spaces of potential interpretations and computing distributions based on interpretations chosen by answerers in the Reddit question thread. |
| Outcome: | The proposed framework analyzes poorly-posed questions using a set of interpretations chosen by human answerers and large language models. |
Understanding Common Ground Misalignment in Goal-Oriented Dialog: A Case-Study with Ubuntu Chat Logs (2025.acl-long)
Copied to clipboard
| Challenge: | a misalignment or misunderstanding can disrupt communication, leading to confusion or conflict. |
| Approach: | They study failures of grounding in Ubuntu IRC datasets to identify misalignments . they find disruptions in conversational flow are driven by a divergence in beliefs . |
| Outcome: | The findings show that misalignment in common ground can disrupt communication . the study also shows that miscommunications can lead to confusion or conflict . |
Addressing Blind Guessing: Calibration of Selection Bias in Multiple-Choice Question Answering by Video Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing MCQA benchmarks fail to capture the full reasoning capabilities of video language models due to selection bias. |
| Approach: | They propose a method to reduce selection bias in video-to-text LLMs by suppressing "blind guessing" they propose 'bold' calibration technique to balance selection bias. |
| Outcome: | The proposed method reduces selection bias and improves model performance compared to existing methods. |
Towards Reward Fairness in RLHF: From a Resource Allocation Perspective (2025.acl-long)
Copied to clipboard
| Challenge: | if rewards are imperfect, they can adversely affect the alignment of large language models (LLMs). |
| Approach: | They propose a bias-agnostic method to address the issue of reward unfairness from a resource allocation perspective without specifically designing for each type of bias . they apply methods Fairness Regularization and Fairness Coefficient to achieve fairness in rewards. |
| Outcome: | The proposed method achieves fairness in rewards while minimizing biases . it can be applied to verification and reinforcement learning scenarios . |
Taming LLMs with Gradient Grouping (2025.acl-long)
Copied to clipboard
| Challenge: | a new study presents scaling with gradient grouping (SGG) the adaptive learning rate scaling approach is based on per-parameter statistics, which incurs memory overhead. |
| Approach: | They propose an optimizer wrapper that improves adaptive learning rate estimation by dynamic grouping and group-specific scaling. |
| Outcome: | The proposed algorithm improves learning rate estimation on diverse models with different model sizes and batch sizes. |
LazyReview: A Dataset for Uncovering Lazy Thinking in NLP Peer Reviews (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models struggle to detect lazy thinking in a zero-shot setting, but instruction-based fine-tuning significantly boosts performance by 10-20 performance points. |
| Approach: | They propose to use LazyReview to train junior reviewers in the community to detect lazy thinking in peer-review sentences annotated with fine-grained lazy thinking categories. |
| Outcome: | The proposed dataset shows that LLMs struggle to detect lazy thinking instances in a zero-shot setting, while instruction-based fine-tuning significantly boosts performance by 10-20 performance points. |
Revisiting Common Assumptions about Arabic Dialects in NLP (2025.acl-long)
Copied to clipboard
| Challenge: | Existing assumptions about Arabic dialect variation are not quantitatively verified. |
| Approach: | They extend and analyze Arabic dialects to assess their validity using a multi-label dataset . they find that the assumptions oversimplify reality and are not always accurate . |
| Outcome: | The proposed methods oversimplify reality and are not always accurate, the authors argue . they show that the proposed assumptions oversimply represent reality and may hinder future work . |
Retrieve to Explain: Evidence-driven Predictions for Explainable Drug Target Identification (2025.acl-long)
Copied to clipboard
| Challenge: | Existing language models lack the capability to quantitatively and faithfully compare answer plausibility in terms of supporting evidence. |
| Approach: | They propose a retrieval-based model that scores and ranks all possible answers to a research question based on evidence retrieved from a document corpus. |
| Outcome: | The proposed model scores and ranks all possible answers to a scientific question based on evidence retrieved from a document corpus . |
Whose Boat Does it Float? Improving Personalization in Preference Tuning via Inferred User Personas (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study shows that LLMs can't tailor outputs to users with uncommon preferences . despite the success of persona inference, we may need debiasing and abstention. |
| Approach: | They propose to use preference data to infer needs and interests of users who prefer either output . they argue that training on preference data augmented with PI boosts personalization . |
| Outcome: | The proposed method can be used to improve personalization with less privacy concerns. |
Which of These Best Describes Multiple Choice Evaluation with LLMs? A) Forced B) Flawed C) Fixable D) All of the Above (2025.acl-long)
Copied to clipboard
| Challenge: | Multiple choice question answering (MCQA) is popular for LLM evaluation due to its simplicity and human-like testing. |
| Approach: | They argue for a reform of multiple choice question answering (MCQA) they argue for more generative formats based on human testing . |
| Outcome: | The proposed reforms improve the quality of MCQA, the authors argue . they show that even when MCQ is a useful format, its datasets suffer from leakage, unanswerability, shortcuts and saturation. |
Detection of Human and Machine-Authored Fake News in Urdu (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fake news detection focus on binary classification and English texts, ignoring the distinction between machine-generated true vs. fake news and low-resource languages. |
| Approach: | They propose to include machine-generated news focusing on Urdu to improve accuracy and robustness. |
| Outcome: | The proposed strategy improves accuracy and robustness across four datasets in various settings. |
An Efficient Task-Oriented Dialogue Policy: Evolutionary Reinforcement Learning Injected by Elite Individuals (2025.acl-long)
Copied to clipboard
| Challenge: | Evolutionary Algorithms (EAs) have been proven to effectively explore the solution space of neural networks by maintaining population diversity. |
| Approach: | They propose an elite individual injection mechanism to enhance EA’s search efficiency by adaptively introducing best-performing individuals into the population. |
| Outcome: | Experiments on four datasets show that the proposed approach significantly improves the balance between exploration and exploitation, boosting performance. |
SR-LLM: Rethinking the Structured Representation in Large Language Model (2025.acl-long)
Copied to clipboard
Jiahuan Zhang, Tianheng Wang, Ziyi Huang, Yulong Wu, Hanqing Wu, DongbaiChen DongbaiChen, Linfeng Song, Yue Zhang, Guozheng Rao, Kaicheng Yu
| Challenge: | Structured representations have long been pivotal in computational linguistics, but their role remains ambiguous in the Large Language Models (LLMs) era. |
| Approach: | They propose a framework that integrates structured representations into LLMs from training-free and training-dependent perspectives. |
| Outcome: | The proposed framework integrates structured representations through natural language descriptions in LLM prompts while augmenting the model’s inference capability through fine-tuning on linguistically described structured representation. |
Taming Language Models for Text-attributed Graph Learning with Decoupled Aggregation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to learning text-attributed graphs neglect interaction between textual and structural information. |
| Approach: | They propose a framework that integrates textual and structural information into TAG learning . they propose combining semantic aggregation and structural aggregations to improve learning a . |
| Outcome: | The proposed framework outperforms state-of-the-art learning methods while requiring less resources. |
Contrastive Prompting Enhances Sentence Embeddings in LLMs through Inference-Time Steering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on prompt engineering to encode the full semantics of a sentence into the embedding of the last token. |
| Approach: | They propose a technique that introduces an extra auxiliary prompt to elicit better sentence embedding . they propose to use the hidden state of the token as the sentence embedded in LLMs . |
| Outcome: | The proposed technique can improve performance of existing prompt-based methods on STS tasks and downstream classification tasks. |
Cracking the Code of Hallucination in LVLMs with Vision-aware Head Divergence (2025.acl-long)
Copied to clipboard
Jinghan He, Kuan Zhu, Haiyun Guo, Junfeng Fang, Zhenglin Hua, Yuheng Jia, Ming Tang, Tat-Seng Chua, Jinqiao Wang
| Challenge: | Existing methods focus on alignment training or decoding refinements but address symptoms at the generation stage without probing the underlying causes. |
| Approach: | They propose a training-free approach to mitigate hallucination by enhancing the role of vision-aware attention heads. |
| Outcome: | The proposed method achieves superior performance compared to state-of-the-art approaches in mitigating hallucinations while maintaining high efficiency with negligible additional time overhead. |
Hierarchical Document Refinement for Long-context Retrieval-augmented Generation (2025.acl-long)
Copied to clipboard
Jiajie Jin, Xiaoxi Li, Guanting Dong, Yuyao Zhang, Yutao Zhu, Yongkang Wu, Zhonghua Li, Ye Qi, Zhicheng Dou
| Challenge: | Real-world RAG applications often encounter long-context input scenarios where redundant information and noise results in higher inference costs and reduced performance. |
| Approach: | They propose an efficient plug-and-play refiner that leverages the structural characteristics of long documents. |
| Outcome: | Experiments on seven QA datasets show that LongRefiner achieves competitive performance in various scenarios while using 10x fewer computational costs and latency compared to baseline. |
Comparing Moral Values in Western English-speaking societies and LLMs with Word Associations (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are trained on extensive corpora to learn linguistic patterns, contextual nuances, and implicit elements of human values. |
| Approach: | They propose to use word associations as low-level underlying representations to obtain a more robust picture of LLMs’ moral reasoning. |
| Outcome: | The proposed method reveals detailed but systematic differences between LLMs and human associations. |
TEACH: A Contrastive Knowledge Adaptive Distillation Framework for Classical Chinese Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Traditional methods for processing classical Chinese segment language understanding into discrete tasks, which overlook crucial background information and reduce user engagement. |
| Approach: | They propose a framework that integrates word sense disambiguation with sentence translation to minimize hallucinations and improve semantic analysis. |
| Outcome: | The proposed framework integrates word sense disambiguation with sentence translation to minimize hallucinations and improve semantic analysis. |
RAG-Critic: Leveraging Automated Critic-Guided Agentic Workflow for Retrieval Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have demonstrated remarkable performance across a wide range of downstream tasks. |
| Approach: | They propose a framework that leverages a critic-guided agentic workflow to improve RAG capabilities autonomously. |
| Outcome: | The proposed framework improves RAG capabilities autonomously by leveraging a critic-guided agentic workflow. |
Progressive Multimodal Reasoning via Active Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to improve multimodal large language models' reasoning performance are limited. |
| Approach: | They propose a framework to progressively improve multimodal reasoning capabilities . they propose active retrieval and Monte Carlo tree search to improve MLLMs' reasoning . |
| Outcome: | The proposed framework improves multimodal reasoning capabilities in multimodal large language models. |
Pre-training Distillation for Large Language Models: A Design Space Exploration (2025.acl-long)
Copied to clipboard
| Challenge: | Knowledge distillation (KD) aims to transfer knowledge from a large teacher model to a smaller student model for model compression. |
| Approach: | They extend knowledge distillation to the pre-training phase of large language models . they first conduct an experiment using a teacher LLM to distill a 1.9B student LLM . |
| Outcome: | The proposed model can be used to distill a 1.9B student model using a teacher LLM. |
Teaching Vision-Language Models to Ask: Resolving Ambiguity in Visual Questions (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research addresses ambiguous visual questions by rephrasing questions, but it fails to address the inherently interactive nature of user interactions with visual language models (VLMs). Existing studies focus on re-phrase questions, and lack of a benchmark to assess VLMs’ capacity for resolving ambiguities through interaction. |
| Approach: | They propose a visual question answering task that provides a natural language answer to a question based on a given image and an automated pipeline to generate ambiguity-clarification question pairs. |
| Outcome: | The proposed benchmark targets three common categories of ambiguity in visual question answering (VQA) context and encompasses various VQA scenarios. |
LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitasks (2025.acl-long)
Copied to clipboard
Yushi Bai, Shangqing Tu, Jiajie Zhang, Hao Peng, Xiaozhi Wang, Xin Lv, Shulin Cao, Jiazheng Xu, Lei Hou, Yuxiao Dong, Jie Tang, Juanzi Li
| Challenge: | Existing benchmarks on longcontext large language models fail to reflect their deep understanding capabilities across diverse tasks. |
| Approach: | They propose a benchmark to assess the ability of long-context large language models to handle long-text problems. |
| Outcome: | The proposed model achieves 50.1% accuracy when directly answering the questions . human experts achieve only 53.7% accuracy under a 15-minute time constraint . |
Battling against Tough Resister: Strategy Planning with Adversarial Game for Non-collaborative Dialogues (2025.acl-long)
Copied to clipboard
| Challenge: | Non-collaborative dialogue involves two participants with conflicting interests engaging in multiround dialogue to achieve their own goals. |
| Approach: | They propose a Game-based Adversarial self-play InterActive training paradigm which constructs an adversarial two-player (a persuader and a resister) zero-sum game and guides the game to approximate Nash Equilibrium (NE) via reinforcement learning. |
| Outcome: | The proposed model achieves state-of-the-art performance on three datasets. |
Cross-model Transferability among Large Language Models on the Platonic Representations of Concepts (2025.acl-long)
Copied to clipboard
| Challenge: | Prior work has shown that a single LLM’s concept representations can be captured as steering vectors (SVs) prior work has demonstrated that SVs extracted from smaller LLMs can effectively control the behavior of larger LLM. |
| Approach: | They propose a linear transformation method to bridge LLM concept representations using simple linear transformations to enable efficient cross-model transfer and behavioral control via SVs. |
| Outcome: | The proposed method bridges concept representations across different LLMs and enables efficient cross-model transfer and behavioral control via SVs. |
FoldMoE: Efficient Long Sequence MoE Training via Attention-MoE Pipelining (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to training LLMs with Mixture-of-Experts (MoE) architecture on long sequences are limited by the insufficient computation. |
| Approach: | They propose a MoE training system that enables token-level overlapping across entire Transformer blocks through novel attention-MoE pipelining. |
| Outcome: | The proposed system achieves 1.49x and 2.72x speedup over state-of-the-art token-level overlapping and non-overlapping baselines on GPT-MoE models with sequences up to 32K tokens. |
LongReward: Improving Long-context Large Language Models with AI Feedback (2025.acl-long)
Copied to clipboard
Jiajie Zhang, Zhongni Hou, Xin Lv, Shulin Cao, Zhenyu Hou, Yilin Niu, Lei Hou, Yuxiao Dong, Ling Feng, Juanzi Li
| Challenge: | In recent years, significant advancements have been achieved in the development of long-context large language models (LLMs). |
| Approach: | They propose a method that utilizes an off-the-shelf LLM to provide rewards for long-context model responses from four human-valued dimensions: helpfulness, logicality, faithfulness, and completeness. |
| Outcome: | The proposed method improves models’ long-context performance and enhances their ability to follow short instructions. |
Influences on LLM Calibration: A Study of Response Agreement, Loss Functions, and Prompt Styles (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies neglect to measure the generalization of their methods to other prompt styles and different sizes of LLMs. |
| Approach: | They propose a framework that trains an auxiliary model for confidence estimation that aggregates responses from multiple LLMs to capture inter-model agreement. |
| Outcome: | The proposed framework integrates response agreement and focal loss with binary cross-entropy to improve calibration from baselines. |
UTBoost: Rigorous Evaluation of Coding Agents on SWE-Bench (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models have enabled the development of coding agents for real-world code generation. |
| Approach: | They propose a novel LLM-driven test case generator that analyzes codebases and dependencies to generate test cases for real-world Python projects. |
| Outcome: | The proposed framework improves the performance of SWE-Bench by analyzing codebases and dependencies. |
Towards Better Evaluation for Generated Patent Claims (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies highlight inconsistencies between automated evaluation metrics and human expert assessments for patent claims. |
| Approach: | They propose a multi-dimensional evaluation method specifically designed for patent claims that incorporates features annotated by patent experts. |
| Outcome: | The proposed method achieves highest correlation with human expert evaluations across all assessment criteria across all tested metrics. |
Fine-Tuning on Diverse Reasoning Chains Drives Within-Inference CoT Refinement in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to generate multiple independent CoTs, combining them through ensembling or other post-hoc strategies, have been shown to be effective in boosting performance. |
| Approach: | They propose a method where LLMs are fine-tuned to generate a sequence of Diverse Chains of Thought (DCoT) within a single inference step. |
| Outcome: | The proposed model can generate multiple chains of thought within a single inference step without external feedback. |
Establishing Trustworthy LLM Evaluation via Shortcut Neuron Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have focused on building dynamic benchmarks to address data contamination issues. |
| Approach: | They propose a method for identifying shortcut neurons through comparative and causal analysis to suppress shortcut neurons. |
| Outcome: | The proposed method overestimates contaminated models and is highly generalizable across benchmarks and hyperparameter settings. |
Do Large Language Models have an English Accent? Evaluating and Improving the Naturalness of Multilingual LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Current Large Language Models (LLMs) are predominantly designed with English as the primary language, but many are still English-dominated. |
| Approach: | They propose to use automatic corpus-level metrics to assess lexical and syntactic naturalness of LLMs in a multilingual context. |
| Outcome: | The proposed method improves naturalness of LLMs in target languages without compromising performance on general-purpose benchmarks. |
Enhancing Character-Level Understanding in LLMs through Token Internal Structure Learning (2025.acl-long)
Copied to clipboard
Zhu Xu, Zhiqiang Zhao, Zihan Zhang, Yuchi Liu, Quanwei Shen, Fei Liu, Yu Kuang, Jian He, Conglin Liu
| Challenge: | Large language models (LLMs) use tokenization methods but often obscure internal character structures within tokens. |
| Approach: | They propose a method that improves models’ ability to capture character positions within tokens by training them on reverse character prediction tasks using the tokenizer’s vocabulary. |
| Outcome: | Experiments show that the proposed method improves position prediction accuracy in large language models, enabling more precise identification of target characters in original text. |
Conformity in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Conformity is a form of social influence that affects the way people respond to information. |
| Approach: | They adapt psychological experiments to examine the extent of conformity in large language models. |
| Outcome: | The proposed interventions mitigate conformity by reducing the naturalness of majority tones and reducing instruction-tuned models. |
Interpret and Improve In-Context Learning via the Lens of Input-Label Mappings (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models excel at downstream NLP tasks through in-context learning . however, the internal mechanisms behind ICL remain under-explored . |
| Approach: | They propose a PC patching approach to identify modules where input-label mappings function . they observe and verify that key heads utilize input-labeled mappings to generate target labels for new queries. |
| Outcome: | The proposed approach detects modules where input-label mappings function . it also detects that key heads use the mappings to generate labels for new queries . |
Positional Overload: Positional Debiasing and Context Window Extension for Large Language Models using Set Encoding (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models typically track the order of tokens using positional encoding, which causes two significant limitations: 1. Positional Bias: When processing long text sequences, the number of token can exceed the range the model was trained on. |
| Approach: | They propose a method that allows multiple pieces of text to be encoded in the same position, eliminating positional bias entirely. |
| Outcome: | The proposed method eliminates positional bias entirely and increases the size of the input an LLM can handle. |
FR-Spec: Accelerating Large-Vocabulary Language Models via Frequency-Ranked Speculative Sampling (2025.acl-long)
Copied to clipboard
Weilin Zhao, Tengyu Pan, Xu Han, Yudi Zhang, Sun Ao, Yuxiang Huang, Kaihuo Zhang, Weilun Zhao, Yuxuan Li, Jie Zhou, Hao Zhou, Jianyong Wang, Maosong Sun, Zhiyuan Liu
| Challenge: | Speculative sampling is an efficient way to accelerate the auto-regressive generation process of large language models. |
| Approach: | They propose a frequency-ranked speculative sampling framework that optimizes draft candidate selection through vocabulary space compression. |
| Outcome: | Experiments show that FR-Spec reduces LM Head computation overhead by 75% while ensuring the equivalence of the final output distribution. |
VReST: Enhancing Reasoning in Large Vision-Language Models through Tree Search and Self-Reward Mechanism (2025.acl-long)
Copied to clipboard
| Challenge: | Large Vision-Language Models (LVLMs) have shown exceptional performance in multimodal tasks, but their effectiveness in complex visual reasoning is constrained. |
| Approach: | They propose a training-free approach that enhances Reasoning in Large Vision-Language Models . they propose integrating Monte Carlo Tree Search and Self-Reward mechanisms into the reasoning tree . |
| Outcome: | The proposed approach surpasses current prompting methods and secures state-of-the-art performance across three multimodal reasoning benchmarks. |
Past Meets Present: Creating Historical Analogy with Large Language Models (2025.acl-long)
Copied to clipboard
Nianqi Li, Siyu Yuan, Jiangjie Chen, Jiaqing Liang, Feng Wei, Zujie Liang, Deqing Yang, Yanghua Xiao
| Challenge: | Historical analogies are important abilities that help people make decisions and understand the world. |
| Approach: | They propose a historical analogy acquisition task that uses large language models to acquire historical analogies. |
| Outcome: | The proposed method mitigates hallucinations and stereotypes when LLMs generate historical analogies. |
Meta-Reflection: A Feedback-Free Reflection Learning Framework (2025.acl-long)
Copied to clipboard
Yaoke Wang, Yun Zhu, XintongBao XintongBao, Wenqiao Zhang, Suyang Dai, Kehan Chen, Wenqiang Li, Gang Huang, Siliang Tang, Yueting Zhuang
| Challenge: | Existing approaches to improve large language models' ability to understand and reason are limited by external feedback. |
| Approach: | They propose a feedback-free reflection mechanism that requires only a single inference pass without external feedback. |
| Outcome: | The proposed method is based on an industrial e-commerce benchmark and public datasets. |
Read it in Two Steps: Translating Extremely Low-Resource Languages with Code-Augmented Grammar Books (2025.acl-long)
Copied to clipboard
| Challenge: | Using code rules improves rule retrieval and application of grammar books in low-resource languages. |
| Approach: | They propose to decompose a grammar rule retrieval and application step into two steps . they propose to represent grammar rules as code functions to facilitate LLM reasoning . |
| Outcome: | The proposed model significantly boosts rule retrieval and application, resulting in 13.1% BLEU improvement. |
Confidence v.s. Critique: A Decomposition of Self-Correction Capability for LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to improve self-correction performance of Large Language Models are based on intrinsic selfcorrectione, which allows the model to check and revise its selfgenerated answers without external feedback. |
| Approach: | They propose to decompose the self-correction capability into confidence and critique capabilities and a metric for overall self-corretion capability evaluation. |
| Outcome: | The proposed method outperforms vanilla SFT and achieves much higher accuracy after self-correction. |
Automating Legal Interpretation with LLMs: Retrieval, Generation, and Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | a novel framework for automated legal interpretation is proposed to alleviate the burden on legal experts. |
| Approach: | They propose a framework for automated legal interpretation that uses large language models to extract concept-related information and interpret legal concepts. |
| Outcome: | The proposed framework eliminates the need for legal experts to interpret legal concepts . it uses large language models to extract concept-related information and interpret legal concept interpretations . |
Visual Evidence Prompting Mitigates Hallucinations in Large Vision-Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | LVLMs have shown impressive progress by integrating visual perception with linguistic understanding to produce contextually grounded outputs. |
| Approach: | They propose a visual evidence prompting method to mitigate hallucinations in large vision-language models by using small visual models to complement them. |
| Outcome: | The proposed method reduces hallucinations by reducing false activation and enhancing correct ones. |
Leveraging Dual Process Theory in Language Agent Framework for Real-time Simultaneous Human-AI Collaboration (2025.acl-long)
Copied to clipboard
Shao Zhang, Xihuai Wang, Wenhao Zhang, Chaoran Li, Junru Song, Tingyu Li, Lin Qiu, Xuezhi Cao, Xunliang Cai, Wen Yao, Weinan Zhang, Xinbing Wang, Ying Wen
| Challenge: | Large language models (LLMs) excel in turn-by-turn human-AI collaboration but struggle with simultaneous tasks requiring real-time interaction. |
| Approach: | They propose a language agent framework that integrates *System 1* and *System 2* for efficient real-time simultaneous human-AI collaboration. |
| Outcome: | The proposed framework improves on existing LLM-based agents and human collaborators by integrating Theory of Mind and asynchronous reflection to infer human intentions and perform reasoning-based autonomous decisions. |
TokAlign: Efficient Vocabulary Adaptation via Token Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Tokenization is a foundational step for Large Language Models (LLMs) but low compression rate of vanilla tokenizers decelerates training and inference process. |
| Approach: | They propose a method to replace the vocabulary of Large Language Models (LLMs) by learning a one-to-one mapping matrix for token IDs. |
| Outcome: | The proposed method significantly improves multilingual text compression rates and vocabulary initialization for Large Language Models. |
AdaEdit: Advancing Continuous Knowledge Editing For Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge editing methods that can efficiently update knowledge in LLMs are limited due to budget constraints. |
| Approach: | They propose a method that can enhance the performance of edited LLMs in large-size continuous editing regimes. |
| Outcome: | Extensive empirical evaluations on multiple LLMs show that the proposed method outperforms existing methods without compromising the general abilities of these models. |
The Impact of Token Granularity on the Predictive Power of Language Model Surprisal (2025.acl-long)
Copied to clipboard
| Challenge: | Word-by-word language model surprisal is often used to model the incremental processing of human readers, but has been overlooked in cognitive modeling due to the granularity of subword tokens. |
| Approach: | They propose to manipulate token granularity to account for processing difficulty of naturalistic text and garden-path constructions. |
| Outcome: | The proposed model can account for the processing difficulty of naturalistic text and garden-path constructions by using tokens defined by a vocabulary size of 8,000. |
Segment-Level Diffusion: A Framework for Controllable Long-Form Generation with Diffusion Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Diffusion models have shown promise in text generation, but often struggle with generating long, coherent, and contextually accurate text. |
| Approach: | They propose a framework that enhances diffusion-based text generation through text segmentation, robust representation training with adversarial and contrastive learning, and improved latent-space guidance. |
| Outcome: | The proposed framework improves diffusion-based text generation and improves scalability and fluency. |
BELLE: A Bi-Level Multi-Agent Reasoning Framework for Multi-Hop Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on multi-hop question answering employ specific methods regardless of question types . complexity of multihop question answerrs often exceeds knowledge boundaries of LLMs . |
| Approach: | They propose a framework that uses chain-of-thought prompting to prompt LLMs to answer multi-hop questions. |
| Outcome: | The proposed framework outperforms baseline models in multi-hop QA scenarios. |
Dynamic and Generalizable Process Reward Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Process Reward Models lack cross-domain generalization and focus on feedback results. |
| Approach: | They propose a process reward model that uses a reward tree to capture and store fine-grained, multi-dimensional reward criteria. |
| Outcome: | The proposed model performs on prevailing benchmarks and out-of-distribution scenarios. |
AdamMeme: Adaptively Probe the Reasoning Capacity of Multimodal Large Language Models on Harmfulness (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models that assess mLLMs on harmful meme understanding are inaccurate and lack accuracy. |
| Approach: | They propose a framework that adaptively probes the reasoning capabilities of mLLMs . their framework systematically reveals the varying performance of different target mllms a . |
| Outcome: | The proposed framework systematically reveals the performance of different target mLLMs. |
Towards Text-Image Interleaved Retrieval (2025.acl-long)
Copied to clipboard
Xin Zhang, Ziqi Dai, Yongqi Li, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Meishan Zhang, Jun Yu, Wenjie Li, Min Zhang
| Challenge: | Existing multimodal information retrieval models rely on single-image inputs . current models use a dense retrieval paradigm, but this approach is not effective . |
| Approach: | They propose a text-image interleaved retrieval task where query and document are interleaves . they adapt off-the-shelf retrievers and build a dense baseline by interleaded multimodal large language model . |
| Outcome: | The proposed model achieves significant improvements over the baseline by substantially fewer visual tokens. |
Large Margin Representation Learning for Robust Cross-lingual Named Entity Recognition (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to name entity recognition neglect distribution skewness and pseudo-label bias . despite promising results, current approaches neglect these problems . |
| Approach: | They propose a framework that optimizes an adaptively reweighted contrastive loss to handle class skewness and pseudo-label bias. |
| Outcome: | The proposed framework outperforms existing methods on multiple benchmarks. |
An Efficient and Precise Training Data Construction Framework for Process-supervised Reward Model in Mathematical Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for constructing process supervision training data are costly or suffer from poor quality. |
| Approach: | They propose a framework called EpicPRM which annotates each intermediate reasoning step based on its quantified contribution and uses an adaptive binary search algorithm to enhance annotation precision and efficiency. |
| Outcome: | The proposed framework improves annotation precision and efficiency and can be used to train a high-quality training dataset with 50k annotated intermediate steps. |
QAEncoder: Towards Aligned Representation Learning in Question Answering Systems (2025.acl-long)
Copied to clipboard
Zhengren Wang, Qinhan Yu, Shida Wei, Zhiyu Li, Feiyu Xiong, Xiaoxing Wang, Simin Niu, Hao Liang, Wentao Zhang
| Challenge: | Modern QA systems entail retrieval-augmented generation (RAG) for accurate and trustworthy responses, but the inherent gap between user queries and relevant documents hinders precise matching. |
| Approach: | They propose a retrieval-augmented generation (RAG)-based approach to bridge this gap by attaching document fingerprints to the embedding to estimate the expectation of potential queries. |
| Outcome: | Experiments across diverse datasets, languages, and embedding models confirm the proposed solution is simple-yet-effective with zero additional index storage, retrieval latency, training costs, or catastrophic forgetting and hallucination issues. |
Game Development as Human-LLM Interaction (2025.acl-long)
Copied to clipboard
| Challenge: | Currently, game development is a highly specialized task that relies on a complex game engine powered by complex programming languages, preventing many gaming enthusiasts from handling it. |
| Approach: | They propose a chat game engine powered by LLM that allows everyone to develop a custom game using natural language through Human-LLM interaction. |
| Outcome: | The proposed engine is designed to support the development of custom games using natural language through Human-LLM interaction. |
Can LLMs Simulate L2-English Dialogue? An Information-Theoretic Analysis of L1-Dependent Biases (2025.acl-long)
Copied to clipboard
Rena Wei Gao, Xuetong Wu, Tatsuki Kuribayashi, Mingrui Ye, Siya Qi, Carsten Roever, Yuanxing Liu, Zheng Yuan, Jey Han Lau
| Challenge: | Large Language Models (LLMs) can simulate non-native-like English use observed in human second language (L2) learners interfered with by their native first language (N1) knowledge. |
| Approach: | They use large language models to simulate non-native-like English use observed in human second language (L2) learners, and then compare their outputs to real L2 learner data. |
| Outcome: | The proposed models replicate L1-dependent patterns observed in human second language (L2) learners, with distinct influences from various languages. |
DeepSolution: Boosting Complex Engineering Solution Design via Tree-based Exploration and Bi-point Thinking (2025.acl-long)
Copied to clipboard
Zhuoqun Li, Haiyang Yu, Xuanang Chen, Hongyu Lin, Yaojie Lu, Fei Huang, Xianpei Han, Yongbin Li, Le Sun
| Challenge: | Existing studies in retrieval-augmented generation (RAG) do not sufficiently address the design of complex engineering solutions. |
| Approach: | They propose a retrieval-augmented generation system that leverages tree-based exploration and bi-point thinking mechanism to generate reliable solutions. |
| Outcome: | Experiments show that the proposed system achieves state-of-the-art (SOTA) performance on the SolutionBench, highlighting its potential to enhance the automation and reliability of complex engineering solution design in real-world applications. |
SurveyPilot: an Agentic Framework for Automated Human Opinion Collection from Social Media (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for opinion survey research exhibit severe biases and lack traceability. |
| Approach: | They propose a finite-state orchestrated agentic framework that automates the collection and analysis of human opinions from social media platforms. |
| Outcome: | The proposed framework achieves close alignment with authentic survey results across multiple domains, with average relative improvements of 68,98% and 51,37% when compared to opinion synthesis and agent-based approaches. |
Sharper and Faster mean Better: Towards More Efficient Vision-Language Model for Hour-scale Long Video Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing multimodal large language models (LLMs) have shown impressive performance on the video understanding task, but extremely long videos still pose significant challenges to their context length, memory consumption, and computational complexity. |
| Approach: | They propose a vision-language model named Sophia for long video understanding which can efficiently handle hour-scale long videos. |
| Outcome: | The proposed model exhibits competitive performance compared to existing video understanding baselines across various benchmarks for long video understanding with reduced time and memory consumption. |
Auto-Arena: Automating LLM Evaluations with Agent Peer Battles and Committee Discussions (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are evolving rapidly and require manual evaluations. |
| Approach: | They propose an LLM-powered framework that automates the entire evaluation process using LLM agents. |
| Outcome: | The proposed framework shows a 92.14% correlation with human preferences, surpassing all previous expert-annotated benchmarks without any manual efforts. |
How Humans and LLMs Organize Conceptual Knowledge: Exploring Subordinate Categories in Italian (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on hierarchical organization of categories focused on basic-1 . but, words at the subordinate level are crucial for effective communication in specialized domains. |
| Approach: | They analyze a psycholinguistic dataset of human-generated exemplars for 187 concrete words . they then evaluate whether textual and vision LLMs produce meaningful exemplar . |
| Outcome: | The results show that human-generated exemplars perform poorly in three key tasks . the results highlight the potential of using AI-generated categories in psycholinguistic research . |
PTQ1.61: Push the Real Limit of Extremely Low-Bit Post-Training Quantization Methods for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for sub 2-bit quantization introduce an extra 1-bit or more per weight. |
| Approach: | They propose a sub 2-bit post-training quantization method that enables weight quantization to 1.61-bit for the first time. |
| Outcome: | The proposed method reduces the upper bound of quantization error to 1.61-bit for the first time. |
ProtoLens: Advancing Prototype Learning for Fine-Grained Interpretability in Text Classification (2025.acl-long)
Copied to clipboard
| Challenge: | ProtoLens provides fine-grained, sub-sentence level interpretability for text classification. |
| Approach: | They propose a prototype-based model that provides fine-grained, sub-sentence level interpretability for text classification. |
| Outcome: | Extensive experiments show that ProtoLens outperforms both prototype-based and non-interpretable baselines on multiple text classification benchmarks. |
Fine-grained Video Dubbing Duration Alignment with Segment Supervised Preference Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Video dubbing systems use neural machine translation and text-to-speech technologies to translate original speech into visual media programs. |
| Approach: | They propose a preference optimization method to optimize video dubbing duration alignment . they propose combining segment-wise sampling and fine-grained loss to mitigate duration mismatches . |
| Outcome: | The proposed method achieves superior performance in duration alignment tasks. |
Sparse Latents Steer Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
Chunlei Xin, Shuheng Zhou, Huijia Zhu, Weiqiang Wang, Xuanang Chen, Xinyan Guan, Yaojie Lu, Hongyu Lin, Xianpei Han, Le Sun
| Challenge: | In this study, we uncover interpretable latents that govern RAG behavior in large language models . Sparse Autoencoders are used to control large language model (LLM) behavior . |
| Approach: | They leverage Sparse Autoencoders within the LLaMA Scope to uncover latents that govern RAG behaviors. |
| Outcome: | The proposed model can be used to control large language models without architectural modifications. |
Unveiling Language-Specific Features in Large Language Models via Sparse Autoencoders (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) exhibit impressive abilities in various domains such as text generation, instruction following, and reasoning. |
| Approach: | They propose a method to decompose the activations of Large Language Models into a sparse linear combination of SAE features. |
| Outcome: | The proposed method shows that some features are strongly related to specific languages, while others are unaffected by ablating them. |
SafeRAG: Benchmarking Security in Retrieval-Augmented Generation of Large Language Model (2025.acl-long)
Copied to clipboard
Xun Liang, Simin Niu, Zhiyu Li, Sensen Zhang, Hanyu Wang, Feiyu Xiong, Zhaoxin Fan, Bo Tang, Jihao Zhao, Jiawei Yang, Shichao Song, Mengwei Wang
| Challenge: | Existing approaches to integrating external knowledge into large language models (LLMs) however, the incorporation of external knowledge increases the vulnerability of LLMs . |
| Approach: | They propose a benchmark to evaluate the RAG security using a dataset . they classify attack tasks into silver noise, inter-context conflict, soft ad, and white Denial-of-Service . |
| Outcome: | The proposed benchmark evaluates the security of RAG against 14 representative RAG components. |
AnRe: Analogical Replay for Temporal Knowledge Graph Forecasting (2025.acl-long)
Copied to clipboard
| Challenge: | Temporal Knowledge Graphs (TKGs) are vital for event prediction, yet current methods face limitations. |
| Approach: | They propose a training-free Analogical Replay reasoning framework that uses LLMs to extract historical contexts and generate analogical reasoning examples as contextual inputs. |
| Outcome: | The proposed model outperforms existing training-free methods on four benchmarks. |
Revisiting the Test-Time Scaling of o1-like Models: Do they Truly Possess Test-Time Scaling Capabilities? (2025.acl-long)
Copied to clipboard
| Challenge: | Longer CoTs of o1-like models do not consistently enhance accuracy, causing performance degradation. |
| Approach: | They propose a method that combines parallel scaling strategies with CoT length characteristics to improve models’ test-time scalability. |
| Outcome: | The proposed method improves models’ test-time scalability compared to majority voting approaches. |
Text is All You Need: LLM-enhanced Incremental Social Event Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing state-of-the-art (SOTA) SED models rely on graph neural networks (GNNs) Existing SED frameworks rely heavily on GNNs, which require complex graph construction and time-consuming training processes. |
| Approach: | They propose a framework that leverages the rich background knowledge of large language models to formalize and disambiguate short texts by completing abbreviations and summarizing informal expressions. |
| Outcome: | The proposed framework outperforms existing models on two challenging real-world datasets. |
Multimodal Pragmatic Jailbreak on Text-to-image Models (2025.acl-long)
Copied to clipboard
Tong Liu, Zhixin Lai, Jiawen Wang, Gengyuan Zhang, Shuo Chen, Philip Torr, Vera Demberg, Volker Tresp, Jindong Gu
| Challenge: | Existing jailbreaks for diffusion-based text-to-image models generate unsafe content . experimental results show that all tested models suffer from unsafe generation . |
| Approach: | They propose a jailbreak that triggers diffusion-based text-to-image models to generate the image with visual text, resulting in unsafe content. |
| Outcome: | The proposed model generates image with visual text, but the model is unsafe under such jailbreak. |
Principled Understanding of Generalization for Generative Transformer Models in Arithmetic Reasoning Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models excel in arithmetic reasoning but their generalization capabilities are incompletely understood. |
| Approach: | They propose a theoretical framework for understanding the generalization behaviors of transformers in arithmetic tasks, focusing on length generalization. |
| Outcome: | The proposed framework can predict generalization behaviors in transformers with a high translation invariance and base mismatch in modular operations. |
Discourse Relation-Enhanced Neural Coherence Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on coherence modeling has focused on integrating entity-based models. |
| Approach: | They propose a model that integrates text- and relation-based features for coherence assessment using position-aware attention and a visible matrix. |
| Outcome: | The proposed model improves baselines on two benchmarks and shows that relation features are important for coherence modeling. |
Benchmarking Open-ended Audio Dialogue Understanding for Large Audio-Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Audio-Language Models (LALMs) have recently unlocked audio dialogue capabilities, enabling direct spoken exchanges with humans. |
| Approach: | They propose to evaluate LALMs' open-ended audio dialogue ability in 3 general scenarios, 12 skills, 9 multilingual languages, and 4 categories of ambiguity handling. |
| Outcome: | The proposed benchmark assesses the open-ended audio dialogue ability for LALMs in 3 general scenarios, 12 skills, 9 multilingual languages, and 4 categories of ambiguity handling. |
from Benign import Toxic: Jailbreaking the Language Model via Adversarial Metaphors (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have exposed the risk of Large Language Models (LLMs) generating harmful content by jailbreak attacks. |
| Approach: | They propose a framework that exploits AdVersArial meTAphoR to induce LLMs to calibrate harmful metaphors for jailbreaking. |
| Outcome: | The proposed framework can successfully jailbreak Large Language Models (LLMs) by leveraging the AdVersArial meTAphoR (AVATAR) framework achieves state-of-the-art attack success rate across multiple advanced LLMs. |
ShifCon: Enhancing Non-Dominant Language Capabilities with a Shift-based Multilingual Contrastive Framework (2025.acl-long)
Copied to clipboard
Hengyuan Zhang, Chenming Shang, Sizhe Wang, Dongdong Zhang, Yiyao Yu, Feng Yao, Renliang Sun, Yujiu Yang, Furu Wei
| Challenge: | Experiments show that ShifCon significantly enhances the performance of non-dominant languages due to the imbalance in training data across languages. |
| Approach: | They propose a Shift-based multilingual Contrastive framework that aligns the internal forward process of other languages toward that of the dominant one. |
| Outcome: | The proposed framework significantly improves performance of non-dominant languages, particularly for low-resource ones. |
MorphMark: Flexible Adaptive Watermarking for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for tracing text origins struggle with a watermark effectiveness dilemma . weaker watermarks preserve text quality, while stronger ones enhance effectiveness . |
| Approach: | They propose a method that adjusts watermark strength in response to changes in a key factor . they first formalize the problem within a multi-objective trade-off analysis framework . |
| Outcome: | The proposed method improves watermark effectiveness but reduces text quality . the proposed method prioritizes flexibility and time and space efficiency . |
A Silver Bullet or a Compromise for Full Attention? A Comprehensive Study of Gist Token-based Context Compression (2025.acl-long)
Copied to clipboard
| Challenge: | gist-based context compression methods can achieve only slight performance loss on tasks like retrieval-augmented generation and long-document QA, but it faces challenges in tasks like synthetic recall. |
| Approach: | They propose two strategies to improve gist-based context compression in large language models. |
| Outcome: | The proposed methods can achieve only slight performance loss on retrieval-augmented generation and long-document QA tasks, but they face challenges in tasks like synthetic recall. |
On the Limit of Language Models as Planning Formalizers (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models can create plans that are neither executable nor verifiable in grounded environments. |
| Approach: | They use Large Language Models to generate a formal representation of the planning domain in some language, such as Planning Domain Definition Language (PDDL). |
| Outcome: | The proposed model outperforms the models directly generating plans while being robust to lexical perturbation. |
Learning to Generate Structured Output with Schema Reinforcement Learning (2025.acl-long)
Copied to clipboard
Yaxi Lu, Haolun Li, Xin Cong, Zhong Zhang, Yesai Wu, Yankai Lin, Zhiyuan Liu, Fangming Liu, Maosong Sun
| Challenge: | Recent advances in large language models have facilitated the development of intelligent applications like automatic web search (Qin et al., 2023) Several methods exist for generating JSON strings from LLMs, including Prompting but often miss certain schemas. |
| Approach: | They propose to use 40K different JSON schemas to assess models' ability to generate valid JSON outputs. |
| Outcome: | The proposed model improves both in generating JSON outputs and downstream tasks. |
Enhancing Unsupervised Sentence Embeddings via Knowledge-Driven Data Augmentation and Gaussian-Decayed Contrastive Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for data augmentation neglect fine-grained knowledge, such as entities and quantities, leading to insufficient diversity and high data noise. |
| Approach: | They propose a pipeline-based data augmentation method via LLMs and introduce the Gaussian-decayed gradient-assisted Contrastive Sentence Embedding (GCSE) model to enhance unsupervised sentence embeddings. |
| Outcome: | The proposed method achieves state-of-the-art performance in semantic textual similarity tasks using fewer data samples and smaller LLMs. |
Improve Safety Training of Large Language Models with Safety-Critical Singular Vectors Localization (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work on safety training with modules such as low-rank adaptation (LoRA) to resist jailbreaks shows promise, but these approaches can inadvertently degrade a model’s general utility. |
| Approach: | They propose a plug-and-play method that locates safety-critical singular vectors within the model's parameter space and a dynamic rank number determination strategy to reduce parameter overhead. |
| Outcome: | The proposed method mitigates the impact of safety training on model utility by explicitly locating and leveraging safety-critical singular vectors within the model’s parameter space. |
WarriorCoder: Learning from Expert Battles to Augment Code Large Language Models (2025.acl-long)
Copied to clipboard
Huawen Feng, Pu Zhao, Qingfeng Sun, Can Xu, Fangkai Yang, Lu Wang, Qianli Ma, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang, Qi Zhang
| Challenge: | Recent code large language models have demonstrated impressive performance on code-related tasks. |
| Approach: | They propose a paradigm that learns from expert battles to address these limitations . they create an arena where leading LLMs challenge each other with evaluations . |
| Outcome: | The proposed model improves on existing models by leveraging expert battles . it achieves state-of-the-art performance even without relying on proprietary models . |
A Triple-View Framework for Fine-Grained Emotion Classification with Clustering-Guided Contrastive Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have focused on dealing with only one of the two difficulties of coarse-grained emotion classification. |
| Approach: | They propose a triple-view framework that treats FEC as an instance-label joint embedding learning problem to tackle both difficulties concurrently by considering three complementary views. |
| Outcome: | The proposed framework achieves significant and consistent improvements on two widely-used benchmark datasets. |
Quantification of Large Language Model Distillation (2025.acl-long)
Copied to clipboard
Sunbowen Lee, Junting Zhou, Chang Ao, Kaige Li, Xeron Du, Sirui He, Haihong Wu, Tianci Liu, Jiaheng Liu, Hamid Alinejad-Rokny, Min Yang, Yitao Liang, Zhoufutu Wen, Shiwen Ni
| Challenge: | Existing studies have revealed the robustness degra-dation caused by data distillation. |
| Approach: | They propose a framework to evaluate and quantify model distillation . they aim to identify identity cognition contradictions and analyse multi-granularity response similarities across models to measure the extent of homogenization. |
| Outcome: | The proposed framework addresses two key aspects: (1) Identifying identity cognition contradictions to assess discrepancies in how models perceive and represent identity-related information; (2) Analyzing multi-granularity response similarities across models to measure the extent of homogenization. |
Demons in the Detail: On Implementing Load Balancing Loss for Training Specialized Mixture-of-Expert Models (2025.acl-long)
Copied to clipboard
Zihan Qiu, Zeyu Huang, Bo Zheng, Kaiyue Wen, Zekun Wang, Rui Men, Ivan Titov, Dayiheng Liu, Jingren Zhou, Junyang Lin
| Challenge: | Existing Mixture-of-Experts training frameworks use a micro-batch to calculate LBL . micro-batches are restricted to a single sequence, preventing expert specialization . |
| Approach: | They propose to use a global-batch to loosen the load balance constraint for MoEs models . they propose to synchronize fi across micro-batches and then use it to calculate the LBL . |
| Outcome: | The proposed global-batch LBL improves the domain specialization of experts . the micro-battery LBL is almost at the sequence level, and the router is pushed to distribute the token evenly . |
Pandora’s Box or Aladdin’s Lamp: A Comprehensive Analysis Revealing the Role of RAG Noise in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Retrieval-Augmented Generation (RAG) has emerged as a promising approach to address hallucinations in large language models (LLMs). |
| Approach: | They define seven distinct noise types from a linguistic perspective and establish a Noise RAG Benchmark (NoiserBench) they propose to evaluate noise that is beneficial to LLMs and noise that's harmful to LRMs. |
| Outcome: | The proposed framework consists of seven distinct noise types from a linguistic perspective and includes multiple datasets and reasoning tasks. |
Stepwise Reasoning Disruption Attack of LLMs (2025.acl-long)
Copied to clipboard
Jingyu Peng, Maolin Wang, Xiangyu Zhao, Kai Zhang, Wanyu Wang, Pengyue Jia, Qidong Liu, Ruocheng Guo, Qi Liu
| Challenge: | Existing attacks on LLM reasoning are constrained by specific settings or lack of imperceptibility, limiting their feasibility and generalizability. |
| Approach: | They propose a stepwise rEasoning error disruption attack that subtly injects errors into prior reasoning steps to mislead the model into producing incorrect subsequent reasoning and final answers. |
| Outcome: | The proposed attack is compatible with zero-shot and few-shot settings, maintains the natural reasoning flow, and ensures covert execution without modification of the instruction. |
Crowd Comparative Reasoning: Unlocking Comprehensive Evaluations for LLM-as-a-Judge (2025.acl-long)
Copied to clipboard
Qiyuan Zhang, Yufei Wang, Yuxin Jiang, Liangyou Li, Chuhan Wu, Yasheng Wang, Xin Jiang, Lifeng Shang, Ruiming Tang, Fuyuan Lyu, Chen Ma
| Challenge: | Existing methods rely on majority voting or criteria expansion to capture detailed and detailed details, often leading to incomplete outcomes. |
| Approach: | They propose a method which introduces additional crowd responses to compare with the candidate responses, thereby exposing deeper and more comprehensive details within the candidate answers. |
| Outcome: | Experiments show that the proposed method improves evaluation reliability and achieves an average gain of 6.7% across five benchmarks. |
Lost in Multilinguality: Dissecting Cross-lingual Factual Inconsistency in Transformer Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Multilingual language models store factual knowledge across languages but struggle to provide consistent responses to semantically equivalent prompts in different languages. |
| Approach: | They propose a linear shortcut method that bypasses computations in the final layers . this method improves accuracy and cross-lingual consistency . |
| Outcome: | The proposed method improves prediction accuracy and cross-lingual consistency. |
Optimizing Decomposition for Optimal Claim Verification (2025.acl-long)
Copied to clipboard
| Challenge: | Existing decomposition and verification paradigms ignore their interactions and potential misalignment. |
| Approach: | They propose a reinforcement learning framework that leverages verifier feedback to learn a policy for dynamically decomposing claims to verifier-preferred atomicity. |
| Outcome: | The proposed framework outperforms existing decomposition policies in verification confidence tests . it improves accuracy and confidence by 0.12 on average across varying verifiers, datasets, and atomcities of input claims. |
GradOT: Training-free Gradient-preserving Offsite-tuning for Large Language Models (2025.acl-long)
Copied to clipboard
Kai Yao, Zhaorui Tan, Penglei Gao, Lichun Li, Kaixin Wu, Yinggui Wang, Yuan Zhao, Yixin Ji, Jianke Zhu, Wei Wang
| Challenge: | Existing methods for offsite-tuning of large language models require high computational costs and lack theoretical analysis. |
| Approach: | They propose an offsite-tuning approach that selectively applies compression techniques such as rank compression and channel pruning to preserve the gradients of fine-tuned adapters while ensuring privacy. |
| Outcome: | The proposed method surpasses existing OT methods in privacy protection and model performance. |
Knowledge Boundary of Large Language Models: A Survey (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) store vast amount of knowledge in their parameters, but they still have limitations in the memorization and utilization of certain knowledge. |
| Approach: | They propose a comprehensive definition of the LLM knowledge boundary and introduce a formalized taxonomy categorizing knowledge into four distinct types. |
| Outcome: | The proposed definition of the LLM knowledge boundary and taxonomy categorizes knowledge into four distinct types . aims to offer a comprehensive overview, facilitate access to key issues, and inspire further advancements in LLM research. |
Mitigating Visual Forgetting via Take-along Visual Conditioning for Multi-modal Long CoT Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) have demonstrated enhanced reasoning capabilities, evolving from simple Chain-of-Thought (CoT) prompting to advanced, product-oriented solutions like OpenAI o1 . |
| Approach: | They propose a strategy that shifts image input to critical reasoning stages and compresses redundant visual tokens via dynamic pruning. |
| Outcome: | The proposed model achieves state-of-the-art on five mathematical reasoning benchmarks (+3.4% vs previous sota) and demonstrates iterative reasoning capabilities for complex multi-step tasks. |
MoC: Mixtures of Text Chunking Learners for Retrieval-Augmented Generation System (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for text chunking are limited by text chunks and lack of domain-specific knowledge. |
| Approach: | They propose a dual-metric evaluation method to quantify text chunking quality . they aim to generate a structured list of chunking regular expressions . |
| Outcome: | The proposed method enables direct quantification of chunking quality . it substantiates the need to integrate LLMs into chunking process . |
Mitigating Selection Bias with Node Pruning and Auxiliary Options (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models exhibit systematic preferences for answer choices when answering multiple-choice questions. |
| Approach: | They propose two methods to identify and remove internal sources of selection bias . they propose Choice Kullback-Leibler Divergence (CKLD) to capture distributional imbalances in model predictions. |
| Outcome: | The proposed methods improve answer accuracy while reducing selection bias. |
Dually Self-Improved Counterfactual Data Augmentation Using Large Language Model (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to generate counterfactual data augmentation are limited due to imbalance and biases in real-world training data. |
| Approach: | They propose a self-improved method for generating high-quality counterfacts using large language models. |
| Outcome: | The proposed method generates high-quality counterfacts on the natural language inference task using lightweight and task-specific LLMs. |
RPO: Retrieval Preference Optimization for Robust Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models struggle to evaluate the correctness of non-parametric knowledge when it differs from internal memorization, leading to knowledge conflicts during response generation. |
| Approach: | They propose a lightweight alignment method to leverage multi-source knowledge based on retrieval relevance. |
| Outcome: | Experiments on four datasets show that the proposed method outperforms RAG by 4-10% in accuracy without any extra component. |
Learning to Reason from Feedback at Test-Time (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to utilizing feedback are expensive and lack the time to perform iterative interactions with the environment. |
| Approach: | They propose a novel paradigm that formulates feedback utilization as an optimization problem at test time and a learnable test-time optimizer to effectively exploit feedback. |
| Outcome: | The proposed paradigm improves scalability and performance on two large language models across four reasoning datasets. |
L-CiteEval: A Suite for Evaluating Fidelity of Long-context Models (2025.acl-long)
Copied to clipboard
| Challenge: | Long-context models (LCMs) have seen remarkable advancements in recent years, facilitating tasks like long-document QA. |
| Approach: | They propose an out-of-the-box suite that can assess both generation quality and fidelity in long-context understanding tasks. |
| Outcome: | The proposed suite can assess both generation quality and fidelity in long-context understanding tasks. |
SECRET: Semi-supervised Clinical Trial Document Similarity Search (2025.acl-long)
Copied to clipboard
| Challenge: | Clinical trials are resource-intensive, time-consuming and expensive to conduct, where errors in trial design, reduced efficacy, and safety events can result in significant delays, financial losses, and damage to reputation. |
| Approach: | They propose a method to summarize clinical trial protocols and search for similar trials based on a query protocol to identify similar historical trials. |
| Outcome: | The proposed method significantly outperforms baselines in recall@1 and precision@1 searches and zero-shot patient-trial matching, highlighting its superior utility in these tasks. |
Geometric Signatures of Compositionality Across a Language Model’s Lifetime (2025.acl-long)
Copied to clipboard
| Challenge: | linguistic compositionality allows atoms to locally combine to create global meaning . a rich array of meanings at the level of a phrase may be explained by simple rules of composition. |
| Approach: | They propose to relate the degree of compositionality in a dataset to the intrinsic dimension of its representations under an LM, a measure of feature complexity. |
| Outcome: | The proposed model is based on a geometric view of the compositionality of a dataset and the intrinsic dimension of its representations under an LM. |
Pattern Recognition or Medical Knowledge? The Problem with Multiple-Choice Questions in Medicine (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are often evaluated using multiple-choice questions (MCQs) modeled on exams like the USMLE. |
| Approach: | They created a fictional medical benchmark centered on an imaginary organ, the Glianorex, to separate memorized knowledge from reasoning ability. |
| Outcome: | The proposed model outperforms base models in English but not in French. |
People who frequently use ChatGPT for writing tasks are accurate and robust detectors of AI-generated text (2025.acl-long)
Copied to clipboard
| Challenge: | Qualitative analysis of experts’ free-form explanations shows that while they rely heavily on specific lexical clues (‘AI vocabulary’), they also pick up on more complex phenomena within the text (e.g., formality, originality, clarity). |
| Approach: | They hire annotators to read 300 non-fiction English articles, label them as either human-written or AI-generated, and provide paragraph-length explanations for their decisions. |
| Outcome: | The annotators who frequently use LLMs for writing tasks outperform commercial and open-source detectors even without evasion tactics like paraphrasing and humanization. |
YuLan-Mini: Pushing the Limits of Open Data-efficient Language Model (2025.acl-long)
Copied to clipboard
Hu Yiwen, Huatong Song, Jie Chen, Jia Deng, Jiapeng Wang, Kun Zhou, Yutao Zhu, Jinhao Jiang, Zican Dong, Yang Lu, Xu Miao, Xin Zhao, Ji-Rong Wen
| Challenge: | prevailing pre-training approaches for large language models involve several complexities. |
| Approach: | They propose a low-cost training recipe and a robust optimization approach to mitigate training instability . they also propose synthesis, curriculum, and data selection pipelines to integrate data . |
| Outcome: | The proposed model achieves top-tier performance among models with similar parameter scale . it is comparable to industry-leading models that require significantly more data . |
Your Model is Overconfident, and Other Lies We Tell Ourselves (2025.acl-long)
Copied to clipboard
| Challenge: | Analyzing 29 models, we find that difficulty is not linear or monotonic. |
| Approach: | They examine the interplay and divergence among various metrics for assessing intrinsic difficulty, including annotator dissensus, training dynamics, and model confidence. |
| Outcome: | The proposed model is based on 29 models on three datasets and analyzed by a linguistics team. |
Bridging the Language Gaps in Large Language Models with Inference-Time Cross-Lingual Intervention (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to address performance gaps in LLMs rely on pretraining or fine-tuning, which are resource-intensive. |
| Approach: | They propose a framework that aligns LLMs' internal representations with those of high-performing languages during inference. |
| Outcome: | The proposed framework improves performance on low-performing (source) languages by aligning their internal representations with those of high-performing languages during inference. |
Plug-in and Fine-tuning: Bridging the Gap between Small Language Models and Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are renowned for their extensive linguistic knowledge and strong generalization capabilities, but their high computational demands make them unsuitable for resource-constrained environments. |
| Approach: | They propose a framework that integrates a single frozen layer from an LLM into a SLM and fine-tunes the combined model for specific tasks. |
| Outcome: | The proposed framework improves performance across a range of natural language processing tasks, including both natural language understanding and generation. |
What is Stigma Attributed to? A Theory-Grounded, Expert-Annotated Interview Corpus for Demystifying Mental-Health Stigma (2025.acl-long)
Copied to clipboard
| Challenge: | Existing resources for training neural models to finely classify mental-health stigma are limited, relying primarily on social media or synthetic data without theoretical underpinnings. |
| Approach: | They propose to use an expert-annotated corpus of human-chatbot interviews to finely classify mental-health stigma. |
| Outcome: | The proposed corpus can facilitate research on computationally detecting, neutralizing, and counteracting mental-health stigma. |
ATRI: Mitigating Multilingual Audio Text Retrieval Inconsistencies by Reducing Data Distribution Errors (2025.acl-long)
Copied to clipboard
Yuguo Yin, Yuxin Xie, Wenyuan Yang, Dongchao Yang, Jinghan Ru, Xianwei Zhuang, Liming Liang, Yuexian Zou
| Challenge: | Existing multilingual audio-text retrieval schemes suffer from inconsistencies for instance similarity matching across languages. |
| Approach: | They propose a multilingual audio-text retrieval scheme that mitigates the impact of data distribution error on recall and consistency. |
| Outcome: | The proposed scheme achieves state-of-the-art performance on recall and consistency metrics for eight mainstream languages, including English. |
Enhancing Transformers for Generalizable First-Order Logical Entailment (2025.acl-long)
Copied to clipboard
Tianshi Zheng, Jiazheng Wang, Zihao Wang, Jiaxin Bai, Hang Yin, Zheye Deng, Yangqiu Song, Jianxin Li
| Challenge: | Moreover, transformers have demonstrated proficiency in logical reasoning over natural language. |
| Approach: | They propose a logic-aware architecture that improves the performance in generalizable first-order logical entailment by combining distribution shifts and unseen knowledge. |
| Outcome: | The proposed architecture outperforms methods designed specifically for knowledge graph query answering on a dataset with a large dataset. |
Self-Taught Agentic Long Context Understanding (2025.acl-long)
Copied to clipboard
Yufan Zhuang, Xiaodong Yu, Jialian Wu, Ximeng Sun, Ze Wang, Jiang Liu, Yusheng Su, Jingbo Shang, Zicheng Liu, Emad Barsoum
| Challenge: | Extensive experiments across seven long-context tasks demonstrate that AgenticLU significantly outperforms state-of-the-art prompting methods and specialized long-consumer LLMs. |
| Approach: | They propose a framework to enhance an LLM's understanding of long-context questions by integrating targeted self-clarification with contextual grounding within an agentic workflow. |
| Outcome: | The proposed framework outperforms state-of-the-art prompting methods and specialized long-context LLMs in seven long-constitut tasks. |
Hallucination Detox: Sensitivity Dropout (SenD) for Large Language Model Training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have focused on identifying and addressing hallucinations in large language models (LLMs), but the impact of the training process on hallucinosity remains underexplored. |
| Approach: | They propose a training protocol to reduce hallucination variance by dropping embedding indices with significant variability and an unsupervised halluciation detection metric, Efficient EigenScore. |
| Outcome: | The proposed training protocol reduces hallucination variance during training by dropping embedding indices with significant variability. |
OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis (2025.acl-long)
Copied to clipboard
Qiushi Sun, Kanzhi Cheng, Zichen Ding, Chuanyang Jin, Yian Wang, Fangzhi Xu, Zhenyu Wu, Chengyou Jia, Liheng Chen, Zhoumianze Liu, Ben Kao, Guohao Li, Junxian He, Yu Qiao, Zhiyong Wu
| Challenge: | Graphical User Interface (GUI) agents powered by Vision-Language Models (VLMs) have demonstrated human-like computer control capability. |
| Approach: | They propose a GUI data synthesis pipeline that reverse engineers GUI trajectory construction process by executing pre-defined tasks. |
| Outcome: | The proposed GUI data synthesis pipeline overcomes the bottlenecks of previous methods that rely on pre-defined tasks and limited data diversity. |
CORAL: Learning Consistent Representations across Multi-step Training with Lighter Speculative Drafter (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods that focus on training and inference suffer from misalignment . speculative decoding is a powerful technique that accelerates large language models . |
| Approach: | They propose a framework that improves both accuracy and efficiency in speculative drafting by using cross-step representational alignment. |
| Outcome: | The proposed framework outperforms existing methods on three LLM families and three benchmark datasets. |
ConSim: Measuring Concept-Based Explanations’ Effectiveness with Automated Simulatability (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation metrics focus only on the quality of the induced space of possible concepts, neglecting the latter. |
| Approach: | They propose to use large language models as simulators to approximate the evaluation and report various analyses to make such approximations reliable. |
| Outcome: | The proposed framework allows for scalable and consistent evaluation across models and datasets. |
Decoding Reading Goals from Eye Movements (2025.acl-long)
Copied to clipboard
| Challenge: | a study examines whether readers can distinguish between two types of reading goals: information seeking and ordinary reading for comprehension. |
| Approach: | They propose a method to distinguish between two types of reading goals: information seeking and ordinary reading for comprehension. |
| Outcome: | The proposed model solves the reading goal-oriented task with the most accurate predictions in real time, the authors say . |
Uncovering Visual-Semantic Psycholinguistic Properties from the Distributional Structure of Text Embedding Space (2025.acl-long)
Copied to clipboard
| Challenge: | Imageability and concreteness are psycholinguistic properties that link visual and semantic spaces. |
| Approach: | They propose an unsupervised measure that quantifies sharpness of peaks in an image-caption dataset. |
| Outcome: | The proposed method is more robust than existing methods and predicts these properties for classification. |
GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent (2025.acl-long)
Copied to clipboard
| Challenge: | GUI automation is a key challenge in dynamic environments. |
| Approach: | They propose a training-free GUI agent that integrates two mechanisms to explore trajectories in GUIs. |
| Outcome: | The proposed GUI-explorer shows significant improvements over existing agents. |
P2 Law: Scaling Law for Post-Training After Model Pruning (2025.acl-long)
Copied to clipboard
| Challenge: | Pruning has become a widely adopted technique for reducing the hardware requirements of large language models (LLMs). |
| Approach: | They propose to use model pruning techniques to maintain high performance while reducing hardware requirements for large language models (LLMs). |
| Outcome: | The proposed model pruning law can be generalized to larger dataset sizes, larger model sizes, and higher pruning rates, offering valuable insights for resource allocation in pruned LLMs. |
Making FETCH! Happen: Finding Emergent Dog Whistles Through Common Habitats (2025.acl-long)
Copied to clipboard
| Challenge: | Dog whistles are coded expressions with dual meanings that slip by content moderation filters . a new study finds that state-of-the-art systems fail to identify novel dog whistles . |
| Approach: | They propose a task to find novel dog whistles in massive social media corpora . they use a strong baseline system that combines vector databases and Large Language Models to identify new dog whistle. |
| Outcome: | The proposed system fails to identify dog whistles across three social media cases . it combines vector databases and Large Language Models to efficiently and effectively identify new dog whistle expressions. |
Lost in the Context: Insufficient and Distracted Attention to Contexts in Preference Modeling (2025.acl-long)
Copied to clipboard
Shihan Dou, Jiayi Chen, Chenhao Huang, Feng Chen, Wei Chengzhi, Huiyuan Zheng, Shichun Liu, Yan Liu, Chenxiao Liu, Chao Xin, Lin Yan, Zongzhang Zhang, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Existing reward models concatenate contexts and responses, but they often ignore crucial segments of the context that are important for evaluating the response quality. |
| Approach: | They propose a reward model that evaluates the response quality based on a given context and assigns a rewards reward. |
| Outcome: | The proposed framework significantly improves preference modeling by increasing attention to relevant information within the context and achieves better generalizability. |
Entailment-Preserving First-order Logic Representations in Natural Language Entailment (2025.acl-long)
Copied to clipboard
| Challenge: | First-order logic (FOL) is often used to represent logical entailment, but determining natural language (NL) enanglement using FOL remains a challenge. |
| Approach: | They propose an Entailment-Preserving FOL representations task and a method which trains an NL-to-FOL translator by using the natural language entailment labels as verifiable rewards. |
| Outcome: | The proposed method achieves 1.8–2.7% improvement in EPR and 17.4–20.6% increase in E PR@16 compared to baselines in three datasets. |
Enhancing Multimodal Continual Instruction Tuning with BranchLoRA (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to fine tune Multimodal Large Language Models (MLLMs) are prone to Catastrophic Forgetting (CF) existing approaches rely on the Mixture-of-Experts (MoE) LoRA framework to preserve previous instruction alignments. |
| Approach: | They propose an asymmetric tuning-freezing mechanism to mitigate parameter inefficiency . branch-specific routers are introduced to ensure optimal branch distribution over time . |
| Outcome: | The proposed framework outperforms existing frameworks on the latest MCIT benchmarks. |
Enhancing Automated Interpretability with Output-Centric Feature Descriptions (2025.acl-long)
Copied to clipboard
| Challenge: | Existing pipelines for describing concepts in large language models (LLMs) rely on activating inputs to generate descriptions that fail to capture the causal effect of a feature on outputs. |
| Approach: | They propose to use inputs that activate features to generate natural language descriptions for concepts represented by large language models (LLMs) . identifying activating inputs is costly, and the mechanistic role of a feature in model behavior is determined by how inputs cause it to activate and how feature activation affects outputs. |
| Outcome: | The proposed methods capture the causal effect of a feature on outputs better than input-centric descriptions, but combine the two leads to the best performance on both input and output evaluations. |
Towards Effective and Efficient Continual Pre-training of Large Language Models (2025.acl-long)
Copied to clipboard
Jie Chen, Zhipeng Chen, Jiapeng Wang, Kun Zhou, Yutao Zhu, Jinhao Jiang, Yingqian Min, Xin Zhao, Zhicheng Dou, Jiaxin Mao, Yankai Lin, Ruihua Song, Jun Xu, Xu Chen, Rui Yan, Zhewei Wei, Di Hu, Wenbing Huang, Ji-Rong Wen
| Challenge: | Continual pre-training (CPT) has been an important approach for adapting language models to specific domains or tasks. |
| Approach: | They propose a Continual pre-training method that can greatly improve Chinese language ability and scientific reasoning ability of LLMs. |
| Outcome: | The proposed method can greatly improve Chinese language ability and scientific reasoning ability of LLMs. |
Efficient Universal Goal Hijacking with Semantics-guided Prompt Organization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for prompt injection have focused on optimizing the suffix, overlooking the role of the prompt. |
| Approach: | They propose a method that incorporates an efficient optimization algorithm and two semantics-guided prompt organization strategies to optimize the suffix sequence for universal goal hijacking. |
| Outcome: | The proposed method can generate a fixed suffix that can concatenate to arbitrary user prompts for universal goal hijacking. |
mPLUG-DocOwl2: High-resolution Compressing for OCR-free Multi-page Document Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal Large Language Models (MLLMs) have improved document understanding performance but generate thousands of visual tokens for a single document image, leading to excessive GPU memory and slower inference times. |
| Approach: | They propose a high-resolution document compression module to generate 324 tokens for a single document image. |
| Outcome: | The proposed module reduces first token latency by more than 50% and improves document comprehension performance. |
What Makes a Good Natural Language Prompt? (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on prompt quality show imbalanced support across models and tasks, and research gaps. |
| Approach: | They propose a property- and human-centric framework for evaluating prompt quality . they propose comparing prompt quality to other factors such as adverbs and apverbs . |
| Outcome: | The proposed framework reveals imbalanced support across models and tasks and substantial research gaps. |
X-TURING: Towards an Enhanced and Efficient Turing Test for Long-Term Dialogue Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Traditional Turing test limits each participant to one message at a time and requires constant human participation. |
| Approach: | They propose to enhance the original Turing test with a burst dialogue pattern, allowing more dynamic exchanges using consecutive messages. |
| Outcome: | The proposed test improves the original test with a burst dialogue pattern, allowing more dynamic exchanges using consecutive messages. |
Are Rules Meant to be Broken? Understanding Multilingual Moral Reasoning as a Computational Pipeline with UniMoral (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to analyze moral reasoning are discordant and lack cohesion, focusing on isolated aspects of the process. |
| Approach: | They propose a unified dataset that integrates moral dilemmas annotated with labels for action choices, ethical principles, contributing factors, and consequences, and captures diverse socio-cultural contexts. |
| Outcome: | The proposed dataset integrates moral dilemmas annotated with labels for action choices, ethical principles, contributing factors, and consequences, along with annotators’ moral and cultural profiles. |
Modality-Aware Neuron Pruning for Unlearning in Multimodal Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models and Multimodal Large Language Modells can memorize sensitive information, raising ethical and privacy concerns. |
| Approach: | They propose a novel unlearning framework that selectively clips neurons based on their relative importance to the targeted forget data. |
| Outcome: | The proposed framework selectively clips neurons based on their relative importance to the targeted forget data, curated for different modalities. |
NGQA: A Nutritional Graph Question Answering Benchmark for Personalized Health-aware Nutritional Reasoning (2025.acl-long)
Copied to clipboard
Zheyuan Zhang, Yiyang Li, Nhi Ha Lan Le, Zehong Wang, Tianyi Ma, Vincent Galassi, Keerthiram Murugesan, Nuno Moniz, Werner Geyer, Nitesh V Chawla, Chuxu Zhang, Yanfang Ye
| Challenge: | Diet plays a critical role in human health, but tailoring dietary reasoning to individual health conditions remains a challenge. |
| Approach: | a new benchmark evaluates dietary reasoning using a national health survey data set. |
| Outcome: | The NGQA benchmark evaluates dietary reasoning across three tasks using a set of question complexity settings and baseline models. |
ReLearn: Unlearning via Learning for Large Language Models (2025.acl-long)
Copied to clipboard
Haoming Xu, Ningyuan Zhao, Liming Yang, Sendong Zhao, Shumin Deng, Mengru Wang, Bryan Hooi, Nay Oo, Huajun Chen, Ningyu Zhang
| Challenge: | Existing methods for unlearning large language models often rely on reverse optimization to reduce target token probabilities. |
| Approach: | They propose a data augmentation and fine-tuning pipeline for effective unlearning . they propose augmentation, evaluation frameworks to measure contextual forgetting . |
| Outcome: | The proposed framework achieves targeted forgetting while preserving high-quality outputs. |
Understanding Cross-Domain Adaptation in Low-Resource Topic Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing topic modeling models struggle in low-resource settings where data is limited . et al., 2003: domain adaptation for low-source topic modeling is challenging in low resources . |
| Approach: | They propose a domain adaptation framework that disentangles domaininvariant and domain-specific components to improve topic adaptation. |
| Outcome: | The proposed model outperforms state-of-the-art methods on low-resource datasets on diverse datasets. |
UAlign: Leveraging Uncertainty Estimations for Factuality Alignment on Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) often struggle to accurately express factual knowledge, especially in cases where the knowledge boundaries are ambiguous. |
| Approach: | They propose a framework that leverages Uncertainty estimations to represent knowledge boundaries and incorporates these representations into prompts for LLMs to Align with factual knowledge. |
| Outcome: | The proposed framework significantly improves the LLMs’ capacities to confidently answer known questions and refuse unknown questions on both in-domain and out-of-domain tasks. |
CoT-Valve: Length-Compressible Chain-of-Thought Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Wei et al., 2022) have developed a powerful method for enhancing the reasoning capabilities of large language models. |
| Approach: | They propose to use a tuning and inference strategy to control the length of reasoning chains by a parameter space direction to control their length. |
| Outcome: | The proposed method reduces reasoning chains on GSM8K from 741 to 225 tokens with a minor performance drop (95.07% to 94.92%) and on AIME from 6827 to 4629 tokens, with only one additional incorrect answer. |
HoH: A Dynamic Benchmark for Evaluating the Impact of Outdated Information on Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Current approaches to addressing knowledge outdating in LLMs struggle with retrieval and generation aspects when handling outdated information. |
| Approach: | They propose a benchmark to evaluate the impact of outdated information on RAG . they use token-level diff algorithms and LLM pipelines to create a large-scale QA dataset . |
| Outcome: | The proposed benchmark analyzes the impact of outdated information on RAG performance. |
Uncertainty Propagation on LLM Agent (2025.acl-long)
Copied to clipboard
Qiwei Zhao, Dong Li, Yanchi Liu, Wei Cheng, Yiyou Sun, Mika Oishi, Takao Osaki, Katsushi Matsuda, Huaxiu Yao, Chen Zhao, Haifeng Chen, Xujiang Zhao
| Challenge: | Existing methods for estimating uncertainty in large language models (LLMs) focus on final-step outputs, which fail to account for cumulative uncertainty over multi-step decision-making process and dynamic interactions between agents and their environments. |
| Approach: | They propose a framework that propagates uncertainty through each step of an LLM-based agent’s reasoning process. |
| Outcome: | Extensive experiments on benchmark datasets show that the proposed framework outperforms state-of-the-art methods by 20%. |
Beyond Position: the emergence of wavelet-like properties in Transformers (2025.acl-long)
Copied to clipboard
| Challenge: | Despite its widespread adoption, theoretical limitations in positional encodings are resolved by developing emergent, wavelet-like processing strategies. |
| Approach: | They propose to use Rotary Position Embeddings to develop emergent, wavelet-like properties that compensate for the positional encoding’s theoretical limitations. |
| Outcome: | The attention heads evolve to implement multi-resolution processing analogous to wavelet transforms. |
Are the Hidden States Hiding Something? Testing the Limits of Factuality-Encoding Capabilities in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies suggest that LLMs encode internal representations of factuality when generating inaccurate or fabricated content. |
| Approach: | They propose a strategy for sampling plausible true-false factoid sentences from tabular data and a procedure for generating realistic, LLM-dependent true-False datasets from Question Answering collections. |
| Outcome: | The proposed approach lays the groundwork for future research on factuality in LLMs and offers practical guidelines for more effective evaluation. |
Disentangling Biased Knowledge from Reasoning in Large Language Models via Machine Unlearning (2025.acl-long)
Copied to clipboard
Zheyuan Liu, Suraj Maharjan, Fanyou Wu, Rahil Parikh, Belhassen Bayar, Srinivasan H. Sengamedu, Meng Jiang
| Challenge: | Existing approaches to disentangle biased knowledge from reasoning are sub-optimal . entangled data makes curation difficult, leading to inclusion of sensitive, toxic data. |
| Approach: | They propose a framework that selectively removes biased knowledge while preserving reasoning abilities. |
| Outcome: | The proposed framework improves fairness accuracy by 14.7% and reasoning performance by 62.6% across multiple LLMs. |
LLaMAs Have Feelings Too: Unveiling Sentiment and Emotion Representations in LLaMA Models Through Probing (2025.acl-long)
Copied to clipboard
Dario Di Palma, Alessandro De Bellis, Giovanni Servedio, Vito Walter Anelli, Fedelucio Narducci, Tommaso Di Noia
| Challenge: | Large Language Models (LLMs) have become central to NLP, demonstrating their ability to adapt to various tasks through prompting techniques. |
| Approach: | They probe the hidden layers of Large Language Models to identify where sentiment features are most represented and to assess how this affects sentiment analysis. |
| Outcome: | The proposed approach enables sentiment tasks to be performed with memory requirements reduced by an average of 57%. |
CxGGEC: Construction-Guided Grammatical Error Correction (2025.acl-long)
Copied to clipboard
| Challenge: | Current GEC methods rely on grammatical labels for syntactic information, often overlooking the inherent usage patterns of language. |
| Approach: | They propose to use construction grammar to capture underlying language patterns and guide corrections by decoding construction tokens into their original forms and correcting erroneous tokens. |
| Outcome: | The proposed model captures underlying language patterns and corrects erroneous construction tokens on English and Chinese benchmarks. |
Beyond Sequences: Two-dimensional Representation and Dependency Encoding for Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing code generation approaches represent code as a linear sequence of tokens, but positional encodings compromise generalization . explicit positional encoders sacrifice permutation invariance, imposes a strict order on the input sequence . |
| Approach: | They propose to represent code snippets as two-dimensional entities with explicit encodings . they propose to use dictionary learning to perform semantic matching between code lines . |
| Outcome: | The proposed model captures the hierarchical and spatial structure of code, especially the dependencies between code lines. |
HD-NDEs: Neural Differential Equations for Hallucination Detection in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Hallucination is a significant challenge for large language models, but current methods struggle when non-factual information arises in the early or mid-sequence of outputs, reducing their reliability. |
| Approach: | They propose a method that captures the full dynamics of large language models by using neural differential equations to assess the truthfulness of statements. |
| Outcome: | The proposed method achieves 14% improvement in AUC-ROC on the True-False dataset compared to state-of-the-art methods. |
What Is That Talk About? A Video-to-Text Summarization Dataset for Scientific Presentations (2025.acl-long)
Copied to clipboard
Dongqi Liu, Chenxi Whitehouse, Xi Yu, Louis Mahon, Rohit Saxena, Zheng Zhao, Yifu Qiu, Mirella Lapata, Vera Demberg
| Challenge: | VISTA dataset contains 18,599 recorded AI conference presentations . large multimodal models exhibit reduced performance in scientific contexts, study shows . |
| Approach: | They propose a dataset specifically designed for video-to-text summarization in scientific domains. |
| Outcome: | This paper compares the performance of large models with human models and shows that they improve on human models. |
NeuSym-RAG: Hybrid Neural Symbolic Retrieval with Multiview Structuring for PDF Question Answering (2025.acl-long)
Copied to clipboard
Ruisheng Cao, Hanchong Zhang, Tiancheng Huang, Zhangyi Kang, Yuxin Zhang, Liangtai Sun, Hanqi Li, Yuxun Miao, Shuai Fan, Lu Chen, Kai Yu
| Challenge: | Existing approaches to retrieval augmented generation neglect PDF structure and layout . individual PDFs often exceed prompt limits and user queries may span multiple documents. |
| Approach: | They propose a hybrid neural symbolic retrieval framework which combines both paradigms in an interactive process. |
| Outcome: | The proposed framework organizes semi-structured PDF content into relational database and vectorstore . it defeats both RAG and structured baselines on three PDF-based QA datasets . |
ProvBench: A Benchmark of Legal Provision Recommendation for Contract Auto-Reviewing (2025.acl-long)
Copied to clipboard
Xiuxuan Shen, Zhongyuan Jiang, Junsan Zhang, Junxiao Han, Yao Wan, Chengjie Guo, Bingcheng Liu, Jie Wu, Renxiang Li, Philip S. Yu
| Challenge: | Contract review is labor-intensive, time-consuming, and costly . a benchmark is proposed to detect potential legal conflicts . |
| Approach: | They propose a benchmark for legal provision recommendation and conflict detection for contract auto-reviewing which aims to recommend the legal provisions related to contract clauses and detect possible legal conflicts. |
| Outcome: | The proposed task recommends legal provisions related to contract clauses and detects legal conflicts. |
F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching (2025.acl-long)
Copied to clipboard
| Challenge: | Recent research in Text-to-Speech (TTS) has experienced great advancement . current models can synthesize speech for any given text and mimic the speaker of audio prompt. |
| Approach: | They propose a fully non-autoregressive text-to-speech system based on flow matching with Diffusion Transformer (DiT) without complex designs such as duration model, text encoder, and phoneme alignment, the text input is simply padded with filler tokens to the same length as input speech, and then denoising is performed for speech generation. |
| Outcome: | The proposed system achieves an inference RTF of 0.15, which is greatly improved compared to state-of-the-art diffusion-based models. |
AutoMedEval: Harnessing Language Models for Automatic Medical Capability Evaluation (2025.acl-long)
Copied to clipboard
Xiechi Zhang, Zetian Ouyang, Linlin Wang, Gerard De Melo, Zhu Cao, Xiaoling Wang, Ya Zhang, Yanfeng Wang, Liang He
| Challenge: | Existing evaluation methods based on large language models (LLMs) are expensive and lack expertise due to limitations in human expertise. |
| Approach: | They propose an open-source automatic evaluation model with 13B parameters specifically engineered to measure the question-answering proficiency of medical LLMs. |
| Outcome: | The proposed model surpasses baselines in terms of correlation with human judgments. |
CoT-based Synthesizer: Enhancing LLM Performance through Answer Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing inference scaling methods rely heavily on the quality of candidate responses . however, they are unable to produce correct answers when all candidates are flawed . |
| Approach: | They propose a CoT-based inference scaling strategy that leverages CoT reasoning to synthesize superior answers by analyzing complementary information from multiple candidate responses. |
| Outcome: | The proposed method improves performance on four benchmark datasets with seven policy models. |
Efficiently Identifying Watermarked Segments in Mixed-Source Texts (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for watermarking entire documents neglect identifying individual watermark segments within long, mixed-source documents. |
| Approach: | They propose a framework for partial watermark detection that detects whether there is a watermark segment in long text and an adaptive online learning algorithm to pinpoint the precise location of watermark segments. |
| Outcome: | The proposed framework outperforms existing methods and is adaptable to other watermarking techniques. |
Assessing Dialect Fairness and Robustness of Large Language Models in Reasoning Tasks (2025.acl-long)
Copied to clipboard
Fangru Lin, Shaoguang Mao, Emanuele La Malfa, Valentin Hofmann, Adrian de Wynter, Xun Wang, Si-Qing Chen, Michael J. Wooldridge, Janet B. Pierrehumbert, Furu Wei
| Challenge: | a study aims to assess the fairness and robustness of Large Language Models in dialectal queries . speakers of "non-standard" dialects are known to experience implicit and explicit discrimination . |
| Approach: | They propose to use a benchmark to assess the fairness of large language models in dialects . they hire speakers with computer science backgrounds to rewrite seven popular benchmarks based on AAVE . |
| Outcome: | The proposed benchmarks show that most models show significant brittleness and unfairness to queries in AAVE. |
Towards a More Generalized Approach in Open Relation Extraction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing OpenRE methods assume unlabeled data is a mixture of known and novel instances. |
| Approach: | They propose a generalized OpenRE setting that considers unlabeled data as a mixture of known and novel instances. |
| Outcome: | The proposed framework outperforms baselines in relation classification and clustering on three benchmark datasets. |
Adaptive Retrieval Without Self-Knowledge? Bringing Uncertainty Back Home (2025.acl-long)
Copied to clipboard
Viktor Moskvoretskii, Maria Marina, Mikhail Salnikov, Nikolay Ivanov, Sergey Pletenev, Daria Galimzianova, Nikita Krayko, Vasily Konovalov, Irina Nikishina, Alexander Panchenko
| Challenge: | Recent adaptive retrieval methods integrate LLMs’ intrinsic knowledge with external information appealing to LLM self-knowledge, but they often neglect efficiency evaluations and comparisons with uncertainty estimation techniques. |
| Approach: | They propose to integrate LLMs’ intrinsic knowledge with external information appealing to LLM self-knowledge but neglect efficiency evaluations and comparisons with uncertainty estimation techniques. |
| Outcome: | The proposed methods outperform complex pipelines in terms of efficiency and self-knowledge while maintaining comparable QA performance. |
Evaluating Language Models as Synthetic Data Generators (2025.acl-long)
Copied to clipboard
Seungone Kim, Juyoung Suk, Xiang Yue, Vijay Viswanathan, Seongyun Lee, Yizhong Wang, Kiril Gashteovski, Carolin Lawrence, Sean Welleck, Graham Neubig
| Challenge: | Prior studies have focused on developing effective data generation methods, but lack systematic comparison of different LMs as data generators in a unified setting. |
| Approach: | They propose to use a benchmark to compare language models' data generation abilities against a set of standardized settings and metrics. |
| Outcome: | The proposed benchmark provides standardized settings and metrics to evaluate LMs’ data generation abilities. |
Can Graph Descriptive Order Affect Solving Graph Problems with LLMs? (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have achieved significant success in reasoning tasks, including mathematical reasoning and logical deduction. |
| Approach: | They conduct the first comprehensive analysis of how the order of graph descriptions impacts LLM performance. |
| Outcome: | The results show that graph descriptions significantly improve LLMs’ comprehension of graph structures, and the robustness of LLM models to graph description order varies across different tasks. |
Learning to Rewrite: Generalized LLM-Generated Text Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing detectors for Large Language Models (LLMs) struggle to generalize in open-world settings. |
| Approach: | They propose a framework to detect LLM-generated text with exceptional generalization to unseen domains by reinforcing LLMs’ inherent rewriting tendencies. |
| Outcome: | The proposed framework outperforms state-of-the-art detection methods by 23.04% in AUROC, 35.10% for out-of distribution tests, and 48.66% under adversarial attacks. |
Evaluating Multimodal Large Language Models on Video Captioning via Monte Carlo Tree Search (2025.acl-long)
Copied to clipboard
Linhao Yu, Xingguang Ji, Yahui Liu, Fanheng Kong, Chenxi Sun, Jingyuan Zhang, Hongzhi Zhang, V. W., Fuzheng Zhang, Deyi Xiong
| Challenge: | Existing benchmarks and evaluation protocols suffer from inadequate or homogeneous creation of key points, exorbitant cost of data creation, and limited evaluation scopes. |
| Approach: | They propose an automatic framework which leverages Monte Carlo Tree Search to construct numerous and diverse descriptive sentences that thoroughly represent video content in an iterative way. |
| Outcome: | The proposed framework improves MCTS-VCB and DREAM-1K on video captioning tasks by 25.0% and 16.3% respectively. |
GIFT-SW: Gaussian noise Injected Fine-Tuning of Salient Weights for LLMs (2025.acl-long)
Copied to clipboard
Maxim Zhelnin, Viktor Moskvoretskii, Egor Shvetsov, Maria Krylova, Venediktov Egor, Zuev Aleksandr, Evgeny Burnaev
| Challenge: | Recent studies show that a small subset of weights significantly impacts performance. |
| Approach: | They propose a Gaussian noise-injected fine-tuning method that updates only salient weights while injecting Gausssian into non-salient weight. |
| Outcome: | The proposed method outperforms full fine-tuning and PEFT methods under the same computational budget. |
Quaff: Quantized Parameter-Efficient Fine-Tuning under Outlier Spatial Stability Hypothesis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for quantized fine-tuning fail to address activation outliers . existing methods incur high computational/memory costs or fail to adequately address outlier activation . |
| Approach: | They propose a Quantized parameter-efficient fine-tuning framework that suppresses outliers exclusively in invariant channels using lightweight operations. |
| Outcome: | The proposed framework reduces outliers in invariant channels while reducing quantization errors. |
Unsolvable Problem Detection: Robust Understanding Evaluation for Large Multimodal Models (2025.acl-long)
Copied to clipboard
Atsuyuki Miyai, Jingkang Yang, Jingyang Zhang, Yifei Ming, Qing Yu, Go Irie, Yixuan Li, Hai Helen Li, Ziwei Liu, Kiyoharu Aizawa
| Challenge: | Multiple-choice question answering (MCQA) is widely used to assess the understanding capability of Large Multimodal Models (LMMs). |
| Approach: | They propose a task to evaluate the robust understanding capability of Large Multimodal Models (LMMs) they introduce a benchmark to assess performance across various ability dimensions . |
| Outcome: | The proposed model can withhold answers when encountering unsolvable problems of MCQA, proving it understands the answer. |
AlignMMBench: Evaluating Chinese Multimodal Alignment in Large Vision-Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks focus on basic abilities using nonverbal methods, such as yes-no and multiple-choice questions. |
| Approach: | They propose a benchmark that provides more nuanced evaluations of alignment capabilities for large Vision-Language Models (VLMs) they use a rule-calibrated evaluator that exceeds GPT-4's evaluation ability and a “alignment score” to assess the robustness and stability of models across diverse prompts. |
| Outcome: | The proposed benchmark covers 13 tasks across three categories and includes both single-turn and multi-turn dialogue scenarios. |
Biased LLMs can Influence Political Decision-Making (2025.acl-long)
Copied to clipboard
Jillian Fisher, Shangbin Feng, Robert Aron, Thomas Richardson, Yejin Choi, Daniel W Fisher, Jennifer Pan, Yulia Tsvetkov, Katharina Reinecke
| Challenge: | Recent studies have found that biased LLMs can influence decisions in areas such as medical classifications and educational hiring. |
| Approach: | They conducted two interactive experiments on partisan bias in large language models while completing tasks with either a biased liberal, biased conservative, or unbiased control model. |
| Outcome: | The results show that prior knowledge of AI is weakly correlated with a reduction of the bias, suggesting that AI education can be crucial for mitigating bias effects. |
LexTempus: Enhancing Temporal Generalizability of Legal Language Models Through Dynamic Mixture of Experts (2025.acl-long)
Copied to clipboard
| Challenge: | a rapid evolution of legal concepts requires that legal language models adapt swiftly accounting for the temporal dynamics. |
| Approach: | They propose a dynamic mixture of experts model that explicitly models the temporal evolution of legal language in an online learning framework. |
| Outcome: | The proposed model can model the temporal evolution of legal language without forgetting past knowledge. |
That is Unacceptable: the Moral Foundations of Canceling (2025.acl-long)
Copied to clipboard
| Challenge: | Annotators' canceling attitudes are influenced by the type of controversial events and involved celebrities. |
| Approach: | They propose to annotate canceling incidents from YouTube and an annotated corpus of videos that are based on their morality to determine their canceling attitudes. |
| Outcome: | The dataset analyzes canceling attitudes of annotators from six videos and comments gathered from YouTube. |
FloorPlan-LLaMa: Aligning Architects’ Feedback and Domain Knowledge in Architectural Floor Plan Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods for floor plan generation rely on statistical metrics like FID, GED, and PSNR, which fail to evaluate using domain knowledge. |
| Approach: | They propose to use a first floor plan dataset to train a floor plan generation model based on a multi-dimensional preference score and a textual analysis to integrate architects’ professional expertise and preferences. |
| Outcome: | The proposed model outperforms baseline models in text-conditional and class-condition tasks and is more rational and aligns better with human preferences. |
TheoremExplainAgent: Towards Video-based Multimodal Explanations for LLM Theorem Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Understanding domain-specific theorems requires more than text-based reasoning . current evaluations of theoretical models are based on textual cues . |
| Approach: | They propose an agentic approach for generating long-form theorem explanation videos using Manim animations. |
| Outcome: | The proposed agent generates long-form theorem explanation videos using Manim animations . the agent achieves a success rate of 93.8% and an overall score of 0.77 . |
FineReason: Evaluating and Improving LLMs’ Deliberate Reasoning through Reflective Puzzle Solving (2025.acl-long)
Copied to clipboard
Guizhen Chen, Weiwen Xu, Hao Zhang, Hou Pong Chan, Chaoqun Liu, Lidong Bing, Deli Zhao, Anh Tuan Luu, Yu Rong
| Challenge: | Recent advances in large language models (LLMs) highlight an important shift from the “System 1” way of quick reactions to the “system 2” style of reflection-and-correction problem solving. |
| Approach: | They propose a logic-puzzle benchmark for systematic evaluation of large language models' reasoning capabilities that decomposes each puzzle into atomic steps. |
| Outcome: | The proposed model improves on state checking and state transition tasks and demonstrates gains in reasoning by up to 5.1%. |
The TIP of the Iceberg: Revealing a Hidden Class of Task-in-Prompt Adversarial Attacks on LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | cipher decoding, riddles, code execution embedded into model prompts bypass safety safeguards of large language models (LLMs) . |
| Approach: | They introduce a novel class of adversarial jailbreak adversarials on large language models, termed Task-in-Prompt (TIP) attacks. |
| Outcome: | The proposed techniques circumvent safeguards in six state-of-the-art language models, including GPT-4o and LLaMA 3.2, and consistently generate restricted content . |
Identifying Reliable Evaluation Metrics for Scientific Text Revision (2025.acl-long)
Copied to clipboard
| Challenge: | Effective revision is a critical step in scientific writing, ensuring clarity, coherence, and adherence to academic standards. |
| Approach: | They propose to use ROUGE and BERTScore to assess revision quality . they also examine LLM-as-a-judge approaches to assess instruction-following revisions . |
| Outcome: | The proposed method improves the accuracy of revision tasks with and without a gold reference. |
Can Language Models Reason about Individualistic Human Values and Preferences? (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods and evaluation frameworks for achieving pluralistic alignment are limited by the diversity of people, which is pre-specified and coarsely categorized, papering over individuality. |
| Approach: | They propose to use a dataset transformed from the influential World Values Survey to study language models on the specific challenge of individualistic value reasoning. |
| Outcome: | The proposed model can predict individualistic values with accuracies between 55% and 65%, while a precise description of individualistic value judgments cannot be approximated only via demographic information. |
BERT-like Models for Slavic Morpheme Segmentation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing morpheme segmentation algorithms for Slavic languages have been improved but performance is still low for words with roots not present in training data. |
| Approach: | They propose to fine-tune BERT-like models for morpheme segmentation using data from Belarusian, Czech, and Russian to account for word semantics. |
| Outcome: | The proposed models outperform all previous approaches in Czech and Russian, with word-level accuracy of 92.5-95.1%. |
Turning Trash into Treasure: Accelerating Inference of Large Language Models with Token Recycling (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) generate only one token at each decoding step, leading to high latency. |
| Approach: | They propose a speculative decoding paradigm that stores tokens in an adjacency matrix and employs a breadth-first-search algorithm to construct a draft tree. |
| Outcome: | The proposed method outperforms existing train-free methods by 30% and even a training method by 25%. |
Unlocking General Long Chain-of-Thought Reasoning Capabilities of Large Language Models via Representation Engineering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work finds that long CoT reasoning can be efficiently elicited by tuning on only a few examples and can easily transfer to other tasks. |
| Approach: | They propose a representation engineering method to unleash the general long CoT reasoning capabilities of LLMs. |
| Outcome: | The proposed method is effective in in-domain and cross-domain scenarios. |
Drift: Enhancing LLM Faithfulness in Rationale Generation via Dual-Reward Probabilistic Inference (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to improving LLM faithfulness rely on superficial calibration methods or costly retraining. |
| Approach: | They propose a probabilistic inference paradigm that leverages task-specific and lookahead rewards to ensure that LLM-generated rationales are more faithful to model decisions. |
| Outcome: | The proposed model improves both accuracy and faithfulness of Large Language Models (LLMs) on three reasoning tasks. |
Fairness through Difference Awareness: Measuring Desired Group Discrimination in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Algorithmic fairness has traditionally adopted the mathematically convenient perspective of racial color-blindness. |
| Approach: | They propose a benchmark suite of eight different scenarios to assess group difference awareness. |
| Outcome: | The proposed model demonstrates that group difference awareness is a distinct dimension to fairness where existing bias mitigation strategies may backfire. |
MergePrint: Merge-Resistant Fingerprints for Robust Black-box Ownership Verification of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Model merging introduces a novel risk of unauthorized use of large language models due to the high cost of training. |
| Approach: | They propose a model merging method that embeds robust fingerprints into models . they aim to protect LLMs from misappropriation via model merg and model theft . |
| Outcome: | The proposed method enables black-box ownership verification without accessing model weights or intermediate outputs. |
Dynamic Scaling of Unit Tests for Code Reward Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing large language models struggle to produce accurate responses on the first attempt for complex reasoning tasks like code generation. |
| Approach: | They propose a lightweight yet effective unit test generator that scales unit tests based on problem difficulty. |
| Outcome: | The proposed approach significantly improves performance on three benchmarks. |
UniConv: Unifying Retrieval and Response Generation for Large Language Models in Conversations (2025.acl-long)
Copied to clipboard
Fengran Mo, Yifan Gao, Chuan Meng, Xin Liu, Zhuofeng Wu, Kelong Mao, Zhengyang Wang, Pei Chen, Zheng Li, Xian Li, Bing Yin, Meng Jiang
| Challenge: | Existing conversational search systems are usually built with two different models . this separation restricts the system from leveraging the model's intrinsic knowledge simultaneously . Existing studies for developing unified models cannot fully address the aspects of understanding conversational context, managing retrieval independently, and generating responses. |
| Approach: | They propose to unify dense retrieval and response generation for large language models in conversation by fine-tuning and mitigating data discrepancy. |
| Outcome: | The proposed model can outperform existing models on five conversational search datasets and reduce inconsistency risks while mitigating data discrepancy. |
Tracking Life’s Ups and Downs: Mining Life Events from Social Media Posts for Mental Health Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have indicated that major life events can greatly impact individuals’ mental health, but shedding its light on social media data is challenging due to the complexity and ambiguity nature of life events. |
| Approach: | They propose to extract life events mentioned in posts on social media to uncover a social media event dataset which includes 12 major life event categories that are likely to occur in everyday life. |
| Outcome: | The proposed dataset includes 12 life event categories that are likely to occur in everyday life and is human-annotated under iterative procedure and boasts a high level of quality. |
ControlSpeech: Towards Simultaneous and Independent Zero-shot Speaker Cloning and Zero-shot Language Style Control (2025.acl-long)
Copied to clipboard
Shengpeng Ji, Qian Chen, Wen Wang, Jialong Zuo, Minghui Fang, Ziyue Jiang, Hai Huang, Zehan Wang, Xize Cheng, Siqi Zheng, Zhou Zhao
| Challenge: | Prior zero-shot TTS models only mimic the speaker’s voice without further control and adjustment capabilities while prior controllable TTS systems cannot perform speaker-specific voice generation. |
| Approach: | They propose a style control module that captures codec representations corresponding to timbre, content, and style in a discrete decoupling codec space. |
| Outcome: | The proposed system can fully clone the speaker's voice and perform speech-specific adjustment and control functions. |
PIC: Unlocking Long-Form Text Generation Capabilities of Large Language Models via Position ID Compression (2025.acl-long)
Copied to clipboard
| Challenge: | Long-context understanding is crucial for large language models (LLMs) however, the ability to “output-long” is underexplored. |
| Approach: | They propose a position ID compression approach to unlock the long-form text generation potential of large language models (LLMs). |
| Outcome: | The proposed approach can extend LLMs' generation length by 1.5 times without compromising generation quality. |
Towards Effective Extraction and Evaluation of Factual Claims (2025.acl-long)
Copied to clipboard
| Challenge: | Lack of a standardized evaluation framework impedes assessment and comparison of claim extraction methods. |
| Approach: | They propose a framework for evaluating claim extraction in the context of fact-checking . they also introduce Claimify, an LLM-based claim extraction method . |
| Outcome: | The proposed evaluation framework outperforms existing methods in the evaluation of claim extraction methods. |
Beyond Facts: Evaluating Intent Hallucination in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) produce unsatisfactory results when faced with complex queries containing multiple conditions. |
| Approach: | They propose a benchmark for intent hallucination that covers 20,068 problems and an automatic LLM generation evaluation metric for detecting intent hallucinosis. |
| Outcome: | The proposed benchmark covers query-only and retrieval-augmented generation (RAG) setups with varying topics and difficulty. |
A Systematic Study of Compositional Syntactic Transformer Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Syntactic language models (SLMs) incorporate syntactical biases into Transformers . authors identify key aspects of design choices in existing models and novel variants based on experimental results . |
| Approach: | They propose a framework that incorporates existing and new SLMs to enhance Transformers by incorporating syntactic biases. |
| Outcome: | The proposed framework improves on existing models and novel variants across language modeling, syntactic generalization, summarization, and inference efficiency. |
M-MAD: Multidimensional Multi-Agent Debate for Advanced Machine Translation Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have shown their potential to deliver human-like judgments. |
| Approach: | They propose a systematic LLM-based multi-agent framework for advanced LLM as-a-judge MT evaluation that integrates dimension-specific results into a final evaluation judgment. |
| Outcome: | The proposed framework outperforms existing LLM-as-a-judge methods and competes with state-of-the-art automatic metrics even when powered by a suboptimal model like GPT-4o mini. |
SongComposer: A Large Language Model for Lyric and Melody Generation in Song Composition (2025.acl-long)
Copied to clipboard
Shuangrui Ding, Zihan Liu, Xiaoyi Dong, Pan Zhang, Rui Qian, Junhao Huang, Conghui He, Dahua Lin, Jiaqi Wang
| Challenge: | Creating lyrics and melodies in symbolic format requires expert knowledge of melody and an advanced understanding of lyrics. |
| Approach: | They introduce SongComposer, a music-specialized large language model that can create symbolic lyrics and melodies following instructions. |
| Outcome: | The proposed model outperforms existing models in symbolic song composition tasks. |
Personalized Text Generation with Contrastive Activation Steering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to personalized text generation rely on retrieval-augmented generation and parameter-efficient fine-tuning. |
| Approach: | They propose a training-free framework that disentangles and represents personalized writing style as a vector in LLM’s activation-space. |
| Outcome: | The proposed framework achieves 8% relative improvement in personalized generation while reducing storage requirements by 1700 over PEFT method. |
Gumbel Reranking: Differentiable End-to-End Reranker Optimization (2025.acl-long)
Copied to clipboard
Siyuan Huang, Zhiyuan Ma, Jintao Du, Changhua Meng, Weiqiang Wang, Jingwen Leng, Minyi Guo, Zhouhan Lin
| Challenge: | Existing distillation-based approaches suffer from training-inference misalignment and fail to capture interdependencies among candidate documents. |
| Approach: | They propose a method to optimize rerankers by learning a stochastic, document-wise Top-k attention mask using the Gumbel Trick and Relaxed Top-K Sampling. |
| Outcome: | The proposed framework minimizes the overall language loss and improves recall on hotpotQA. |
Hybrid Preferences: Learning to Route Instances for Human vs. AI Feedback (2025.acl-long)
Copied to clipboard
Lester James Validad Miranda, Yizhong Wang, Yanai Elazar, Sachin Kumar, Valentina Pyatkin, Faeze Brahman, Noah A. Smith, Hannaneh Hajishirzi, Pradeep Dasigi
| Challenge: | Learning from human feedback has enabled the alignment of language models (LMs) with human preferences. |
| Approach: | They propose a Hybrid Preference routER that defers an annotation to either humans or LMs, achieving better annotation quality while reducing the cost of human-only annotation. |
| Outcome: | The proposed model achieves better annotation quality while reducing the cost of human-only annotation. |
SEOE: A Scalable and Reliable Semantic Evaluation Framework for Open Domain Event Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods for Open Domain Event Detection (ODED) lack representative representations of the real world, making it difficult to accurately reflect performance of various ODED methods in real-world scenarios. |
| Approach: | They propose a scalable and reliable Semantic-level Evaluation framework for Open domain event detection by constructing a more representative evaluation benchmark and introducing a semantic evaluation metric. |
| Outcome: | The proposed framework first constructs a more representative evaluation benchmark that currently includes 564 event types covering 7 major domains, with a cost-effective supplementary annotation strategy to ensure the benchmark’s representativeness. |
The UD-NewsCrawl Treebank: Reflections and Challenges from a Large-scale Tagalog Syntactic Annotation Project (2025.acl-long)
Copied to clipboard
| Challenge: | UD-NewsCrawl is the largest Tagalog treebank to date, with 15.6k trees manually annotated according to the Universal Dependencies framework. |
| Approach: | They propose to use UD-NewsCrawl to annotate Tagalog trees using the Universal Dependencies framework. |
| Outcome: | The proposed treebanks are based on the Universal Dependencies framework and have 15.6k trees annotated manually. |
DRAG: Distilling RAG for SLMs from LLMs to Transfer Knowledge and Mitigate Hallucination via Evidence and Graph-based Distillation (2025.acl-long)
Copied to clipboard
Jennifer Chen, Aidar Myrzakhan, Yaxin Luo, Hassaan Muhammad Khan, Sondos Mahmoud Bsharat, Zhiqiang Shen
| Challenge: | Large-scale RAG systems consume significant computational resources and are prone to generating “hallucinated” content from Humans. |
| Approach: | They propose a framework for distilling RAG knowledge from large-scale language models into small LMs. |
| Outcome: | The proposed method outperforms the prior competitive RAG methods like MiniRAG for SLMs by up to 27.7% using the same models, preserving high-level efficiency and reliability. |
G-Safeguard: A Topology-Guided Security Lens and Treatment on LLM-based Multi-agent Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Model (LLM)-based Multi-agent Systems (MAS) have demonstrated remarkable capabilities in various complex tasks, but their vulnerability to adversarial attacks, misinformation propagation, and unintended behaviors have raised significant concerns. |
| Approach: | They propose a topology-guided security lens and treatment for robust LLM-MAS that leverages graph neural networks to detect anomalies on the multi-agent utterance graph and employ topological intervention for attack remediation. |
| Outcome: | Experiments show that the proposed security lens recovers 40% of the performance under various attack strategies and integrates with mainstream MAS with security guarantees. |
Deontological Keyword Bias: The Impact of Modal Expressions on Normative Judgments of Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are increasingly engaging in moral and ethical reasoning, where criteria for judgment are often unclear, even for humans. |
| Approach: | They propose a judgment strategy that integrates few-shot examples with reasoning prompts to mitigate this bias. |
| Outcome: | The proposed judgment strategy integrates few-shot examples with reasoning prompts to mitigate this bias. |
LegalReasoner: Step-wised Verification-Correction for Legal Judgment Reasoning (2025.acl-long)
Copied to clipboard
Weijie Shi, Han Zhu, Jiaming Ji, Mengze Li, Jipeng Zhang, Ruiyuan Zhang, Jia Zhu, Jiajie Xu, Sirui Han, Yike Guo
| Challenge: | Existing legal judgment prediction methods struggle with logical errors when conducting complex legal reasoning. |
| Approach: | They propose a method which enhances LJP reliability through step-wise verification and correction of the reasoning process. |
| Outcome: | The proposed model significantly improves concordance with court decisions from 72.37 to 80.27 on LLAMA-3.1-70B. |
Rolling the DICE on Idiomaticity: How LLMs Fail to Grasp Context (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models fail to resolve idiomaticity when it depends on contextual understanding . idiom frequency influences performance but does not guarantee accurate interpretation. |
| Approach: | They propose a novel contrastive dataset to assess whether large language models can effectively leverage context to disambiguate idiomatic meanings. |
| Outcome: | The proposed model performs better on sentences deemed more likely by the model . collocational frequency and sentence probability influence performance but not accuracy . |
ChartCoder: Advancing Multimodal Large Language Model for Chart-to-Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing open-source MLLMs fail to fully capture dense information embedded in charts . current models still face significant challenges in understanding and analyzing visual tasks such as captioning and question answering. |
| Approach: | They propose a chart-to-code MLLM which leverages Code LLMs as the language backbone to enhance the executability of the generated code. |
| Outcome: | The proposed model surpasses existing open-source models on chart-to-code benchmarks with only 7B parameters and provides lossless representations that contain all critical details. |
The Cross-linguistic Role of Animacy in Grammar Structures (2025.acl-long)
Copied to clipboard
| Challenge: | Animacy is a semantic feature of nominals and follows a hierarchy of personal pronouns . it is argued that soft tendencies may be the cause of animacy constraints . however, there is no empirical evidence for this . |
| Approach: | They propose a method to reliably classify animacy classes of nominals in 11 languages from 5 families using multilingual large language models and word sense disambiguation datasets. |
| Outcome: | The proposed method shows that animacy displays consistent cross-linguistic tendencies in terms of preferred morphosyntactic constructions, but not always in line with received wisdom. |
LexGen: Domain-aware Multilingual Lexicon Generation (2025.acl-long)
Copied to clipboard
Ayush Maheshwari, Atul Kumar Singh, N J Karthika, Krishnakant Bhatt, Preethi Jyothi, Ganesh Ramakrishnan
| Challenge: | Lexicon generation is a key task in specialized domains due to infrequent usage of terms . a new model is proposed to generate dictionary words for 6 Indian languages . |
| Approach: | They propose a model to generate dictionary words for 6 Indian languages in the multi-domain setting. |
| Outcome: | The proposed model generalizes to unseen domains and unsealed languages. |
How to Train Long-Context Language Models (Effectively) (2025.acl-long)
Copied to clipboard
| Challenge: | a new study shows that language models can process extremely long contexts with minimal training. |
| Approach: | They use supervised fine-tuning and continued training to evaluate a language model's long-context capabilities. |
| Outcome: | The proposed model outperforms Llama-3.1-8B-Instruct on most long-context tasks . the model can process 512K tokens, one of the longest context windows of LMs . |
MathFusion: Enhancing Mathematical Problem-solving of LLM through Instruction Fusion (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown impressive progress in mathematical problem-solving . current approaches to enhance mathematical reasoning focus on instance-level modifications . |
| Approach: | They propose a framework that enhances mathematical reasoning through cross-problem instruction synthesis. |
| Outcome: | The proposed framework boosts mathematical reasoning by 18.0 points while maintaining high data efficiency. |
Mining Complex Patterns of Argumentative Reasoning in Natural Language Dialogue (2025.acl-long)
Copied to clipboard
| Challenge: | Argumentation scheme mining is the task of automatically identifying reasoning mechanisms behind argument inferences. |
| Approach: | They propose to create a corpus of 441 arguments annotated with 24 argumentation schemes and leverage the capabilities of LLMs and Transformer-based models to validate their applicability in real-world scenarios. |
| Outcome: | The proposed corpus of arguments is pre-trained on a large corpus containing textbook-like argumentation schemes and validates their applicability in real-world scenarios. |
OS Agents: A Survey on MLLM-based Agents for Computer, Phone and Browser Use (2025.acl-long)
Copied to clipboard
Xueyu Hu, Tao Xiong, Biao Yi, Zishu Wei, Ruixuan Xiao, Yurun Chen, Jiasheng Ye, Meiling Tao, Xiangxin Zhou, Ziyu Zhao, Yuhuai Li, Shengze Xu, Shenzhi Wang, Xinchen Xu, Shuofei Qiao, Zhaokai Wang, Kun Kuang, Tieyong Zeng, Liang Wang, Jiwei Li, Yuchen Eleanor Jiang, Wangchunshu Zhou, Guoyin Wang, Keting Yin, Zhou Zhao, Hongxia Yang, Fan Wu, Shengyu Zhang, Fei Wu
| Challenge: | a new generation of (M)LLMs is enabling the creation of superintelligent AI assistants . OS Agents can complete tasks autonomously and have the potential to significantly enhance the lives of billions of users worldwide. |
| Approach: | They propose to build OS Agents that operate within operating systems' GUIs and GUIs . they examine evaluation metrics and benchmarks to identify promising directions . |
| Outcome: | The proposed agents are based on operating systems (OS) and operating systems frameworks. |
Data Quality Issues in Multilingual Speech Datasets: The Need for Sociolinguistic Awareness and Proactive Language Planning (2025.acl-long)
Copied to clipboard
| Challenge: | Despite their growing importance, the quality of these datasets remains under-researched. |
| Approach: | They propose guidelines and recommendations to address quality issues in future dataset development . they find that macro-level issues are more prevalent in less institutionalized, often under-resourced languages . |
| Outcome: | The results highlight the need for proactive language planning and enhanced data quality control in the process of automatic speech recognition dataset creation. |
LLM as a Broken Telephone: Iterative Generation Distorts Information (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models are increasingly responsible for online content, but they can be distorted by repeated transmission. |
| Approach: | They investigate whether large language models distort information through iterative generation. |
| Outcome: | The findings raise important questions about the reliability of LLM-generated content in iterative workflows. |
VLM2-Bench: A Closer Look at How Well VLMs Implicitly Link Explicit Matching Visual Cues (2025.acl-long)
Copied to clipboard
| Challenge: | Existing vision-language models lack the ability to visually link matching visual cues across images or frames. |
| Approach: | They propose a benchmark to assess whether vision-language models can Visually Link Matching cues with 9 subtasks and over 3,000 test cases. |
| Outcome: | The proposed benchmarks on multiple images and videos do not demonstrate that vision-language models can link visual cues across images or frames. |
Alleviating Distribution Shift in Synthetic Data for Machine Translation Quality Estimation (2025.acl-long)
Copied to clipboard
| Challenge: | Quality Estimation (QE) models evaluate the quality of machine translations without reference translations, serving as the reward models for the translation task. |
| Approach: | They propose a framework for alleviating distribution shift in synthetic QE data . they employ a constrained beam search algorithm and distinct generation models to enhance translation diversity. |
| Outcome: | The proposed framework outperforms SOTA baselines like CometKiwi in supervised and unsupervised settings. |
Evaluation Agent: Efficient and Promptable Evaluation Framework for Visual Generative Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods rely on rigid pipelines that overlook user needs and provide numerical results without clear explanations. |
| Approach: | They propose an evaluation framework that employs human-like strategies for efficient, dynamic, multi-round evaluations using only a few samples per round. |
| Outcome: | The evaluation agent framework reduces evaluation time to 10% of traditional methods while delivering comparable results. |
Large Language Models Struggle to Describe the Haystack without Human Help: A Social Science-Inspired Evaluation of Topic Models (2025.acl-long)
Copied to clipboard
Zongxia Li, Lorena Calvo-Bartolomé, Alexander Miserlis Hoyle, Paiheng Xu, Daniel Kofi Stephens, Juan Francisco Fung, Alden Dima, Jordan Lee Boyd-Graber
| Challenge: | a common use of NLP is to facilitate the understanding of large document collections. |
| Approach: | They propose to use large language models to replace probabilistic topic models in real-world applications. |
| Outcome: | The proposed model generates more human-readable topics and shows higher average win probabilities than traditional models for data exploration. |
ActiView: Evaluating Active Perception Ability for Multimodal Large Language Models (2025.acl-long)
Copied to clipboard
Ziyue Wang, Chi Chen, Fuwen Luo, Yurui Dong, Yuanchi Zhang, Yuzhuang Xu, Xiaolong Wang, Peng Li, Yang Liu
| Challenge: | Existing benchmarks for evaluating MLLMs have not addressed active perception . a novel benchmark is proposed to evaluate active perception in ML models . |
| Approach: | They propose a benchmark to evaluate active perception in Multimodal Large Language Models . they restrict the perceptual field of a model and require it to actively zoom or shift it . |
| Outcome: | The proposed benchmark focuses on a specialized form of Visual Question Answering (VQA) that eases and quantifies the evaluation yet challenging for existing MLLMs. |
Enough Coin Flips Can Make LLMs Act Bayesian (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models exhibit the ability to generalize given few-shot examples in their input prompt, an emergent capability known as in-context learning. |
| Approach: | They investigate whether large language models use in-context learning to generalize given few-shot examples in their input prompt. |
| Outcome: | The proposed model can generalize given few-shot examples in their input prompt, an emergent capability known as in-context learning. |
GAMEBoT: Transparent Assessment of LLM Reasoning in Games (2025.acl-long)
Copied to clipboard
| Challenge: | Existing efforts to create benchmarks that move beyond superficial pattern recognition to delve into the profound reasoning skills required for problemsolving face challenges such as insufficient interpretability, performance saturation or data contamination. |
| Approach: | They propose a gaming arena designed for rigorous assessment of LLM reasoning capabilities. |
| Outcome: | The proposed framework decomposes complex reasoning into predefined modular subproblems and generates ground truth for these subproblem types. |
A Text is Worth Several Tokens: Text Embedding from LLMs Secretly Aligns Well with The Key Tokens (2025.acl-long)
Copied to clipboard
| Challenge: | et al., 2023) show that text embeddings from large language models can be aligned with key tokens in input text. |
| Approach: | They propose a sparse retrieval method based on aligned tokens for large language models . they show that this phenomenon is universal and is not affected by model architecture . |
| Outcome: | The proposed method can achieve 80% of the dense retrieval effect of the same model while reducing the computation significantly. |
Commonsense Reasoning in Arab Culture (2025.acl-long)
Copied to clipboard
Abdelrahman Sadallah, Junior Cedric Tonga, Khalid Almubarak, Saeed Almheiri, Farah Atif, Chatrine Qwaider, Karima Kadaoui, Sara Shatnawi, Yaser Alesh, Fajri Koto
| Challenge: | Existing studies on commonsense reasoning in Arabic have relied on machine translations that lack cultural depth and introduce anglocentric biases. |
| Approach: | They propose a commonsense reasoning dataset in Arabic that covers 13 Arab countries. |
| Outcome: | The proposed dataset covers 13 countries across the Gulf, Levant, North Africa, and the Nile Valley. |
AXIS: Efficient Human-Agent-Computer Interaction with API-First LLM-Based Agents (2025.acl-long)
Copied to clipboard
Junting Lu, Zhiyang Zhang, Fangkai Yang, Jue Zhang, Lu Wang, Chao Du, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang, Qi Zhang
| Challenge: | Multimodal large language models (MLLMs) have enabled LLM-based agents to directly interact with application user interfaces (UIs), enhancing agents’ performance in complex tasks. |
| Approach: | They propose a novel agent framework that prioritizes actions through application programming interfaces over UI actions and facilitates the creation and expansion of APIs through automated exploration of applications. |
| Outcome: | The proposed framework reduces task completion time by 65%-70% and cognitive workload by 38%-53% while maintaining accuracy of 97%-98% compared to humans. |
Translation and Fusion Improves Cross-lingual Information Extraction (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have shown significant progress in information extraction tasks due to lack of labeled data for fine-tuning and unlabeled text for pre-training. |
| Approach: | They propose a framework in which large language models are fine-tuned to use English translations of low-resource language data. |
| Outcome: | The proposed model improves cross-lingual transfer over the base model on 12 multilingual IE datasets spanning 50 languages. |
Conditional Dichotomy Quantification via Geometric Embedding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods that rely on semantic similarity fail to capture the nuanced oppositional dynamics essential for these applications. |
| Approach: | They propose a task that formalizes the measurement of conditional dichotomy by using a dichotomian framework. |
| Outcome: | The proposed framework provides carefully constructed datasets covering debate, defeasible inference, and causal reasoning scenarios. |
Aligning Large Language Models with Implicit Preferences from User-Generated Content (2025.acl-long)
Copied to clipboard
Zhaoxuan Tan, Zheng Li, Tianyi Liu, Haodong Wang, Hyokun Yun, Ming Zeng, Pei Chen, Zhihan Zhang, Yifan Gao, Ruijie Wang, Priyanka Nigam, Bing Yin, Meng Jiang
| Challenge: | Existing preference learning methods rely heavily on curated data from humans or advanced LLMs, which is costly and difficult to scale. |
| Approach: | They propose a framework that leverages implicit preferences in unlabeled user-generated content to generate preference data. |
| Outcome: | The proposed framework transforms user-generated content into user queries and generates responses from the policy model. |
VQAGuider: Guiding Multimodal Large Language Models to Answer Complex Video Questions (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal large language models (MLLMs) can grasp the intention of a question and decomposing it to a series of visual recognition sub-tasks to find out the answer with the help of an agent. |
| Approach: | They propose a framework for multimodal large language models to grasp the intention of a question and decompose it into a series of visual recognition sub-tasks to find out the answer. |
| Outcome: | The proposed framework improves the accuracy of complex video-related questions by 29.6% and 17.2% on CVQA and the existing VQA datasets. |
Large Language Models are Good Relational Learners (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to serialize large language models disregard critical relational structures and creates redundancies. |
| Approach: | They propose a graph neural network encoder to create structured relational prompts for large language models within a retrieval-augmented generation framework. |
| Outcome: | The proposed architecture preserves relational structure of databases while enabling LLMs to process and reason over complex entity relationships. |
SpaRE: Enhancing Spatial Reasoning in Vision-Language Models with Synthetic Data (2025.acl-long)
Copied to clipboard
| Challenge: | Vision-language models struggle with spatial reasoning, a skill that humans excel at. |
| Approach: | They propose to use a spatial-reasoning Enhanced (SpaRE) VLM to improve spatial reasoning in visual question answering and robotics. |
| Outcome: | The proposed model achieves a 49% performance gain on the What's Up benchmark while maintaining strong results on general tasks. |
Distilling an End-to-End Voice Assistant Without Instruction Training Data (2025.acl-long)
Copied to clipboard
| Challenge: | Recent efforts to train speech-only LLMs have led to models “forging” speech information from text-only models. |
| Approach: | They propose a paradigm for training Speech Large Language Models without instruction data by using the response of a text-only LLM to transcripts as self-supervision. |
| Outcome: | The proposed model generalizes to Spoken Question Answering, Classification, and Translation and achieves a 72% win rate compared with state-of-the-art models like Qwen 2 Audio . |
CoMet: Metaphor-Driven Covert Communication for Multi-Agent Language Games (2025.acl-long)
Copied to clipboard
| Challenge: | Metaphors are crucial for humans to express complex or subtle ideas by comparing one concept to another, often from a different domain. |
| Approach: | They propose a framework that enables LLMs to engage in metaphor processing by combining hypothesis-based metaphor reasoner and metaphor generator. |
| Outcome: | The proposed framework enhances agents' ability to interpret and apply metaphors in language games. |
CER: Confidence Enhanced Reasoning in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to enhance the reliability of Large Language Models (LLMs) in complex reasoning tasks are limited by their limitations. |
| Approach: | They propose an uncertainty-aware framework to enhance the reliability of Large Language Models . they quantify the confidence of intermediate answers and evaluate the reliability based on these confidences a way that reflects the reliability. |
| Outcome: | The proposed approach improves accuracy of large language models in math and open-domain tasks by 7.4% and 5.8% over baseline approaches. |
Watermarking Large Language Models: An Unbiased and Low-risk Method (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have highlighted the risk of misusing them, raising the need for accurate detection of LLM-generated content. |
| Approach: | They propose a method to inject imperceptible identifiers into large language models (LLMs) this method is unbiased and preserves the original token distribution in expectation . |
| Outcome: | The proposed method preserves the original token distribution in expectation and has lower risk of producing unsatisfactory outputs in low-entropy scenarios compared to existing unbiased watermarks. |
On Synthetic Data Strategies for Domain-Specific Generative Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Generative retrieval models can be used to generate ranked lists of potentially relevant document identifiers for a user query. |
| Approach: | They propose a synthetic data generation strategy for a two-stage training framework that focuses on learning to decode document identifiers from queries and a strategy for mining hard negatives based on initial model's predictions. |
| Outcome: | The proposed model can generate ranked lists of potentially relevant document identifiers for a user query and then refine ranking through preference learning. |
LLM Braces: Straightening Out LLM Predictions with Relevant Sub-Updates (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies reveal that much of the knowledge in a Transformer-based Large Language Model (LLM) is encoded in its feed-forward (FFN) layers, where each FNN layer can be interpreted as the summation of sub-updates, each corresponding to a weighted column vector from the FFN’s value parameter matrix. |
| Approach: | They propose a method that computes relevance scores associated with value vectors in FFN layers and leverages these scores to dynamically adjust the contribution of sub-updates. |
| Outcome: | The proposed framework outperforms baseline approaches in fine-tuning and zero-shot settings while requiring significantly fewer tunable parameters. |
CONFETTI: Conversational Function-Calling Evaluation Through Turn-Level Interactions (2025.acl-long)
Copied to clipboard
Tamer Alkhouli, Katerina Margatina, James Gung, Raphael Shu, Claudia Zaghi, Monica Sunkara, Yi Zhang
| Challenge: | Current benchmarks lack comprehensive assessment of large language models in complex conversational scenarios. |
| Approach: | They introduce a conversational benchmark to evaluate function-calling capabilities of large language models. |
| Outcome: | The proposed benchmark evaluates the function-calling capabilities and response quality of large language models. |
Evaluating Theory of (an uncertain) Mind: Predicting the Uncertain Beliefs of Others from Conversational Cues (2025.acl-long)
Copied to clipboard
| Challenge: | Typically, beliefs are held or not held, but there are situations where an individual's beliefs are better represented more flexibly. |
| Approach: | They propose a set of tasks that challenge language models to model the uncertainty of participants in a dialogue. |
| Outcome: | The proposed tasks show that language models can model the uncertainty of participants in a conversation. |
Uncertainty in Causality: A New Frontier (2025.acl-long)
Copied to clipboard
| Challenge: | Existing literature on uncertainty in causality is lacking a comprehensive review of this area. |
| Approach: | They propose a trichotomy categorizing causal uncertainty into aleatoric, epistemic, ontological and ontological categories . they propose key traits for an optimal causal LLM to handle uncertainty . |
| Outcome: | The proposed method categorizes causal uncertainty into aleatoric, epistemic, and ontological uncertainty. |
SynthesizeMe! Inducing Persona-Guided Prompts for Personalized Reward Models in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent calls for pluralistic alignment of Large Language Models encourage adapting models to diverse user preferences. |
| Approach: | They propose a method to induce synthetic user personas from user interactions for personalized reward modeling. |
| Outcome: | The proposed approach improves LLM-as-a-judge accuracy by 4.4% on Chatbot Arena. |
When People are Floods: Analyzing Dehumanizing Metaphors in Immigration Discourse with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | a computational approach to measure metaphorical language is based on immigration discourse on social media. |
| Approach: | They propose a computational approach that leverages word-level and document-level signals to measure metaphor with respect to immigration discourse on social media. |
| Outcome: | The proposed method measures metaphorical language in immigration discourse on social media. |
AGrail: A Lifelong Agent Guardrail with Effective and Adaptive Safety Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing defense agencies fail to adaptively and effectively mitigate these risks. |
| Approach: | They propose a lifelong agent guardrail that enhances LLM agent safety by enabling adaptive safety check generation, effective safety check optimization, and tool compatibility & flexibility. |
| Outcome: | The proposed agent guardrail achieves strong performance against task-specific and systemic risks and is transferable across different LLM agents’ tasks. |
Improving Model Factuality with Fine-grained Critique-based Evaluator (2025.acl-long)
Copied to clipboard
Yiqing Xie, Wenxuan Zhou, Pradyot Prakash, Di Jin, Yuning Mao, Quintin Fettes, Arya Talebzadeh, Sinong Wang, Han Fang, Carolyn Rose, Daniel Fried, Hejia Zhang
| Challenge: | Factuality evaluation aims to detect factual errors produced by language models and guide the development of more factual models. |
| Approach: | They propose a framework that leverages FenCE to improve the factuality of LM generators by constructing training data. |
| Outcome: | The proposed framework improves the factuality of LM generators by enhancing their training data. |
Building a Long Text Privacy Policy Corpus with Multi-Class Labels (2025.acl-long)
Copied to clipboard
| Challenge: | Legal text is susceptible to multiple valid, conflicting interpretations, and indeterminacy, interdependence between clauses, meaningful silence, and implications of legal defaults. |
| Approach: | They propose to annotate privacy policies from 149 firms using a hand-coded dataset that captures key challenges peculiar to legal language. |
| Outcome: | The proposed dataset includes privacy policies from 149 firms and includes materials incorporated by reference. |
R2-MultiOmnia: Leading Multilingual Multimodal Reasoning via Self-Training (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have introduced eclectic strategies to enhance MLLMs’ reasoning capabilities, but they remain related to a single language. |
| Approach: | They propose a modular approach that instructs models to abstract key elements of the reasoning process and refine reasoning trajectories via self-correction. |
| Outcome: | The proposed approach improves multimodal reasoning, gets aligned performances among the languages approaching strong models and improves the model's performance. |
When the LM misunderstood the human chuckled: Analyzing garden path effects in humans and language models (2025.acl-long)
Copied to clipboard
| Challenge: | Modern Large Language Models (LLMs) have shown human-like abilities in many language tasks, sparking interest in comparing LLMs’ and humans’ language processing. |
| Approach: | They propose to answer two questions: 1. What makes garden-path sentences hard for humans? 2. Do the same reasons make garden- path sentences hard? |
| Outcome: | The proposed models show that humans struggle with specific syntactic complexities, with some models showing high correlation with human comprehension. |
Cross-Lingual Pitfalls: Automatic Probing Cross-Lingual Weakness of Multilingual Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models have achieved remarkable success in Natural Language Processing, yet their cross-lingual consistency remains a significant challenge. |
| Approach: | They propose a method to identify cross-lingual weaknesses in Large Language Models . they construct bilingual question pairs that expose performance discrepancies between English and target languages . |
| Outcome: | The proposed method uncovers over 50% accuracy drops in target languages across models. |
VLSBench: Unveiling Visual Leakage in Multimodal Safety (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that textual unlearning does not achieve comparable safety performance with image-text alignment. |
| Approach: | They propose to use textual unlearning to align MLLMs with image-text pairs to explain this problem . they construct a visual leakless safety bench with 2.2k image- text pairs to test this problem. |
| Outcome: | The proposed model can refuse image-text pairs according to textual queries, leading to unreliable safety evaluations. |
Browsing Lost Unformed Recollections: A Benchmark for Tip-of-the-Tongue Search and Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | BLUR is a tip-of-the-tongue known-item search and reasoning benchmark for general AI assistants. |
| Approach: | They introduce a tip-of-the-tongue known-item search and reasoning benchmark for general AI assistants. |
| Outcome: | The proposed benchmark demands searching and reasoning across multimodal and multilingual inputs, as well as proficient tool use, in order to excel on. |
Data Laundering: Artificially Boosting Benchmark Results through Knowledge Distillation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that language model benchmarks are vulnerable to manipulation and exploitation. |
| Approach: | They propose a method that allows the covert transfer of benchmark-specific knowledge through seemingly legitimate intermediate training steps. |
| Outcome: | The proposed method can achieve significant improvements in accuracy without developing reasoning capabilities. |
Conspiracy Theories and Where to Find Them on TikTok (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on TikTok's potential to promote and amplify harmful content have not been conducted. |
| Approach: | They analyze a longitudinal dataset of 1.5M videos shared in the U.S. over three years and evaluate the effects of TikTok’s Creativity Program for monetization. |
| Outcome: | The proposed model achieves high precision in detecting harmful content, but its overall performance is comparable to fine-tuned traditional models such as RoBERTa. |
Growing Through Experience: Scaling Episodic Grounding in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Language models (LMs) require effective episodic grounding to perform well at physical planning tasks due to their limited ability to learn from and apply past experiences. |
| Approach: | They propose a weak-to-strong episodic learning framework that integrates episodic memory into hierarchical representations and pre-trained knowledge to unlock larger LMs' potential for grounding. |
| Outcome: | The proposed framework outperforms top proprietary LMs by 3.45% across diverse planning and question-answering tasks. |
Exploiting the Shadows: Unveiling Privacy Leaks through Lower-Ranked Tokens in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models face vulnerabilities related to the extraction of sensitive information. |
| Approach: | They propose a method to exploit the model's lower-ranked output tokens to extract private information from retrieved documents or training knowledge. |
| Outcome: | The proposed method is effective in both the agentic application privacy extraction setting and the direct training data extraction. |
Attacking Vision-Language Computer Agents via Pop-ups (2025.acl-long)
Copied to clipboard
| Challenge: | Existing tools for analyzing and testing VLMs are lacking in understanding what types of attacks are possible and what types are effective. |
| Approach: | They propose to integrate pop-ups into existing agent testing environments to attack VLM agents by ignoring them. |
| Outcome: | The proposed attack success rate is 86% and decreases by 47% when integrating pop-ups into existing agent testing environments. |
Explicit and Implicit Data Augmentation for Social Event Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Social event detection relies on labeled data, but annotation is costly and labor-intensive. |
| Approach: | They propose a plug-and-play dual augmentation framework that combines explicit text-based and implicit feature-space augmentation to enhance data diversity and model robustness. |
| Outcome: | The proposed framework outperforms the best baseline model by 17.67% on the Twitter2012 dataset and 15.57% on the twitter2018 dataset in terms of the average F1 score. |
In Prospect and Retrospect: Reflective Memory Management for Long-term Personalized Dialogue Agents (2025.acl-long)
Copied to clipboard
Zhen Tan, Jun Yan, I-Hung Hsu, Rujun Han, Zifeng Wang, Long Le, Yiwen Song, Yanfei Chen, Hamid Palangi, George Lee, Anand Rajan Iyer, Tianlong Chen, Huan Liu, Chen-Yu Lee, Tomas Pfister
| Challenge: | Existing approaches to long-term dialogue memory management fail to capture the natural semantic structure of conversations, leading to fragmented and incomplete representations. |
| Approach: | They propose a mechanism that integrates forward- and backward-looking reflections into a personalized memory bank for effective future retrieval. |
| Outcome: | The proposed mechanism outperforms state-of-the-art benchmarks on a long-term dialogue memory model. |
Revisiting Classical Chinese Event Extraction with Ancient Literature Information (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on classical Chinese event extraction focus on grafting the complex modeling from English or modern Chinese works, neglecting the unique characteristic of this language. |
| Approach: | They propose a Literary Vision-Language Model (VLM) for classical Chinese event extraction . they integrate annotations, historical background and character glyphs to capture the inner- and outer-context information from the sequence. |
| Outcome: | The proposed model can capture the inner- and outer-context information at nearly zero cost. |
Unanswerability Evaluation for Retrieval Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation frameworks for retrieval-augmented generation (RAG) systems focus on answerable queries, but ignore the importance of appropriately rejecting unanswerable requests. |
| Approach: | They propose a framework to evaluate whether retrieval-augmented generation systems handle unanswerable queries specific to a given knowledge base. |
| Outcome: | The proposed framework synthesizes diverse and challenging queries for any given knowledge base and evaluates them with unanswered ratio and acceptable ratio metrics. |
SCALE: Towards Collaborative Content Analysis in Social Science with Large Language Model Agents and Human Intervention (2025.acl-long)
Copied to clipboard
| Challenge: | Content analysis is labor-intensive and time-consuming process that requires multiple rounds of manual annotation, domain expert discussion, and rule-based refinement. |
| Approach: | They propose a multi-agent framework that effectively Simulates Content Analysis via Large language model (LLM) ag Ents. |
| Outcome: | The proposed framework achieves human-approximated performance across various content analysis tasks. |
Self-Error-Instruct: Generalizing from Errors for LLMs Mathematical Reasoning (2025.acl-long)
Copied to clipboard
Erxin Yu, Jing Li, Ming Liao, Qi Zhu, Boyang Xue, Minghui Xu, Baojun Wang, Lanqing Hong, Fei Mi, Lifeng Shang
| Challenge: | Existing approaches to learning from errors synthesize training data by extrapolating from isolated bad cases, thereby failing to generalize the extensive patterns inherent within these cases. |
| Approach: | They propose a framework that synthesizes more generalized training data from isolated bad cases by extrapolating from isolated cases. |
| Outcome: | The proposed framework synthesizes more generalized training data to address these model weaknesses. |
RAGEval: Scenario Specific RAG Evaluation Dataset Generation Framework (2025.acl-long)
Copied to clipboard
Kunlun Zhu, Yifan Luo, Dingling Xu, Yukun Yan, Zhenghao Liu, Shi Yu, Ruobing Wang, Shuo Wang, Yishan Li, Nan Zhang, Xu Han, Zhiyuan Liu, Maosong Sun
| Challenge: | Existing evaluation metrics for RAG systems are lacking due to high costs of data construction and lack of factual accuracy. |
| Approach: | They propose a framework to evaluate RAG systems in specialized scenarios . they propose three new metrics to evaluate LLM-generated responses . |
| Outcome: | The proposed framework outperforms zero-shot and one-shot methods in terms of clarity, safety, conformity, and richness of generated samples. |
A Survey on Patent Analysis: From NLP to Multimodal AI (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in pretrained language models and large language models have demonstrated transformative capabilities across diverse domains. |
| Approach: | They propose a taxonomy for categorization based on tasks in the patent life cycle . they introduce a novel taxonomies for categorizing based upon tasks in patent life cycles . |
| Outcome: | The proposed method is based on tasks in the patent life cycle and provides a taxonomy for categorization based upon tasks in patent life cycles. |
SciVer: Evaluating Foundation Models for Multimodal Scientific Claim Verification (2025.acl-long)
Copied to clipboard
| Challenge: | Existing scientific claim verification benchmarks focus on textual content alone or on verifying claims based on a single table. |
| Approach: | They propose to use SciVer to evaluate the ability of foundation models to verify claims within a multimodal scientific context. |
| Outcome: | The proposed model outperforms 21 state-of-the-art models and human experts on SciVer. |
MultiAgentBench : Evaluating the Collaboration and Competition of LLM agents (2025.acl-long)
Copied to clipboard
Kunlun Zhu, Hongyi Du, Zhaochen Hong, Xiaocheng Yang, Shuyi Guo, Zhe Wang, Zhenhailong Wang, Cheng Qian, Robert Tang, Heng Ji, Jiaxuan You
| Challenge: | Large Language Models (LLMs) have shown remarkable capabilities as autonomous agents, yet existing benchmarks focus on single-agent tasks or are confined to narrow domains, failing to capture the dynamics of multi-agent coordination and competition. |
| Approach: | They propose a benchmark to evaluate LLM-based multi-agent systems across diverse, interactive scenarios. |
| Outcome: | The proposed framework measures task completion and quality of collaboration and competition using novel, milestone-based key performance indicators. |
Sinhala Encoder-only Language Models and Evaluation (2025.acl-long)
Copied to clipboard
Tharindu Ranasinghe, Hansi Hettiarachchi, Nadeesha Chathurangi Naradde Vidana Pathirana, Damith Premasiri, Lasitha Uyangodage, Isuri Nanomi Arachchige, Alistair Plum, Paul Rayson, Ruslan Mitkov
| Challenge: | Recent advances in language models (LMs) have produced excellent results in many NLP tasks, but their effectiveness is highly dependent on available pre-training resources. |
| Approach: | They propose to collect the largest monolingual corpus for Sinhala and compile a benchmark and evaluate LMs on it. |
| Outcome: | The proposed language models outperform the popular multilingual LMs in downstream NLP tasks. |
LLMs can Perform Multi-Dimensional Analytic Writing Assessments: A Case Study of L2 Graduate-Level Academic English Writing (2025.acl-long)
Copied to clipboard
| Challenge: | a growing number of studies have indicated the general usefulness of LLMs for automated writing assessments. |
| Approach: | They propose a framework that evaluates LLMs' ability to provide scores and comments based on multiple assessment criteria. |
| Outcome: | The proposed framework is interpretable, cost-efficient, scalable, and reproducible . it is compared to existing methods that rely on manual judgments . |
SEUF: Is Unlearning One Expert Enough for Mixture-of-Experts LLMs? (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in LLMs unlearning have shown remarkable success in removing unwanted data-model influences while preserving the model’s utility for legitimate knowledge. |
| Approach: | They propose a Selected-Expert Unlearning Framework (SEUF) that combines expert attribution and an anchor loss to ensure controlled unlearning. |
| Outcome: | Experiments show that the proposed framework improves forget quality and model utility by 35% on MoE LLMs across benchmarks and LLM architectures compared to standard unlearning algorithms . |
Pragmatics in the Era of Large Language Models: A Survey on Datasets, Evaluation, Opportunities and Challenges (2025.acl-long)
Copied to clipboard
Bolei Ma, Yuting Li, Wei Zhou, Ziwei Gong, Yang Janet Liu, Katja Jasinskaja, Annemarie Friedrich, Julia Hirschberg, Frauke Kreuter, Barbara Plank
| Challenge: | linguistics studies how context influences meaning of language and how people use it to convey implied meanings, emotions, and intentions. |
| Approach: | They analyze task designs, data collection methods, evaluation approaches and their relevance to real-world applications. |
| Outcome: | The findings highlight emerging trends, challenges, and gaps in existing benchmarks . the findings will contribute to more nuanced and context-aware NLP models . |
LocAgent: Graph-Guided LLM Agents for Code Localization (2025.acl-long)
Copied to clipboard
Zhaoling Chen, Robert Tang, Gangda Deng, Fang Wu, Jialong Wu, Zhiwei Jiang, Viktor Prasanna, Arman Cohan, Xingyao Wang
| Challenge: | Existing approaches struggle to efficiently navigate complex codebases when identifying relevant code snippets. |
| Approach: | They propose a graph-guided agent framework that addresses code localization through a distributed graph-based agent. |
| Outcome: | The proposed framework improves accuracy on real-world benchmarks and can be used to locate code snippets at a cost of 86%. |
COSMMIC: Comment-Sensitive Multimodal Multilingual Indian Corpus for Summarization and Headline Generation (2025.acl-long)
Copied to clipboard
Raghvendra Kumar, Mohammed Salman S A, Aryan Sahu, Tridib Nandi, Pragathi Y P, Sriparna Saha, Jose G Moreno
| Challenge: | COSMMIC is a multimodal, multilingual dataset featuring nine major Indian languages. |
| Approach: | They propose a multimodal, multilingual multimodal multimodal dataset that integrates text, images and user feedback to enhance summarization. |
| Outcome: | The proposed dataset is based on 4,959 article-image pairs and 24,484 reader comments with ground-truth summaries available in all included languages. |
Mind the Gap: Static and Interactive Evaluations of Large Audio Models (2025.acl-long)
Copied to clipboard
Minzhi Li, William Barr Held, Michael J Ryan, Kunat Pipatanakul, Potsawee Manakul, Hao Zhu, Diyi Yang
| Challenge: | Recent work has focused on evaluating large audio models (LAMs) that directly accept audio inputs. |
| Approach: | They propose an interactive approach to evaluate large audio models and collect 7,500 LAM interactions from 484 participants. |
| Outcome: | The proposed model is based on a set of user-generated audio interfaces with 7,500 interactions from 484 participants. |
Understanding In-Context Machine Translation for Low-Resource Languages: A Case Study on Manchu (2025.acl-long)
Copied to clipboard
| Challenge: | In-context machine translation (MT) with large language models can take advantage of linguistic resources such as grammar books and dictionaries. |
| Approach: | They propose to use in-context machine translation (MT) with large language models to take advantage of linguistic resources such as grammar books and dictionaries. |
| Outcome: | The proposed approach can take advantage of dictionaries and grammar books, but its performance is poor for many lowresource languages. |
CKnowEdit: A New Chinese Knowledge Editing Dataset for Linguistics, Facts, and Logic Error Correction in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | CKnowEdit is the first-ever knowledge editing dataset designed to correct linguistic, factual, and logical errors in Large Language Models. |
| Approach: | They propose a Chinese knowledge editing dataset to correct linguistic, factual, and logical errors in Large Language Models. |
| Outcome: | The proposed dataset highlights the challenges that LLMs face in mastering Chinese . CKnowEdit can correct linguistic, factual, and logical errors in Chinese, the authors show . |
TripleFact: Defending Data Contamination in the Evaluation of LLM-driven Fake News Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation paradigms for fake news detection are based on static datasets and closed-world assumptions that are inadvertently memorized during pre-training. |
| Approach: | They propose a framework to mitigate BDC risk while prioritizing real-world applicability by integrating three components to assess robustness against human-crafted misinformation. |
| Outcome: | The proposed framework mitigates BDC risk while prioritizing real-world applicability. |
Meaning Beyond Truth Conditions: Evaluating Discourse Level Understanding via Anaphora Accessibility (2025.acl-long)
Copied to clipboard
| Challenge: | Existing assessments of understanding at the lexical and sentence levels are limited to lexica and sentence level, but few of them target whether LLMs accurately represent and update states of natural language discourse. |
| Approach: | They propose anaphora accessibility as a diagnostic for assessing discourse understanding . they use a dataset inspired by theoretical research in dynamic semantics to evaluate human and LLM performance. |
| Outcome: | The proposed dataset shows that humans and LLMs align on some tasks and diverge on others. |
Large Language and Reasoning Models are Shallow Disjunctive Reasoners (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) struggle with systematic reasoning on out-of-distribution (OOD) tasks. |
| Approach: | They propose to use a set of constraints to measure OOD generalization to create large reasoning models that can be leveraged to solve real-world problems. |
| Outcome: | The proposed models outperform their LLM counterparts in single-path reasoning tasks but struggle in multi-path setting. |
Warmup Generations: A Task-Agnostic Approach for Guiding Sequence-to-Sequence Learning with Unsupervised Initial State Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing supervised fine-tuning (SFT) methods focus on directly generating the target output without leveraging the benefits of intermediate steps or initial guidance. |
| Approach: | They propose a task-agnostic framework that enables models to generate intermediate "warmup" sequences that are iteratively refined to maximize their contribution to the final output. |
| Outcome: | The proposed framework outperforms traditional supervised fine-tuning methods on translation, summarization, and multi-choice question answering tasks. |
Building Better: Avoiding Pitfalls in Developing Language Resources when Data is Scarce (2025.acl-long)
Copied to clipboard
| Challenge: | Language is a powerful means of communication and should be regarded as more than just a collection of tokens. |
| Approach: | They collect feedback from individuals directly involved in and impacted by NLP artefacts for medium- and low-resource languages and highlight key issues related to data quality, cultural appropriateness and ethics of common annotation practices. |
| Outcome: | The findings highlight key issues related to data quality, cultural appropriateness, and ethics of common annotation practices. |
BRIGHTER: BRIdging the Gap in Human-Annotated Textual Emotion Recognition Datasets for 28 Languages (2025.acl-long)
Copied to clipboard
Shamsuddeen Hassan Muhammad, Nedjma Ousidhoum, Idris Abdulmumin, Jan Philip Wahle, Terry Ruas, Meriem Beloucif, Christine de Kock, Nirmal Surange, Daniela Teodorescu, Ibrahim Said Ahmad, David Ifeoluwa Adelani, Alham Fikri Aji, Felermino D. M. A. Ali, Ilseyar Alimova, Vladimir Araujo, Nikolay Babakov, Naomi Baes, Ana-Maria Bucur, Andiswa Bukula, Guanqun Cao, Rodrigo Tufiño, Rendi Chevi, Chiamaka Ijeoma Chukwuneke, Alexandra Ciobotaru, Daryna Dementieva, Murja Sani Gadanya, Robert Geislinger, Bela Gipp, Oumaima Hourrane, Oana Ignat, Falalu Ibrahim Lawan, Rooweither Mabuya, Rahmad Mahendra, Vukosi Marivate, Alexander Panchenko, Andrew Piper, Charles Henrique Porto Ferreira, Vitaly Protasov, Samuel Rutunda, Manish Shrivastava, Aura Cristina Udrea, Lilian Diana Awuor Wanzare, Sophie Wu, Florian Valentin Wunderlich, Hanif Muhammad Zhafran, Tianhui Zhang, Yi Zhou, Saif M. Mohammad
| Challenge: | Emotion recognition is an umbrella term for several NLP tasks, but most work on high-resource languages has focused on low-resourced languages. |
| Approach: | They propose to use emotion recognition to describe perceived emotions in 28 different languages and across several domains to identify and annotate the datasets. |
| Outcome: | The proposed datasets cover low-resource languages from Africa, Asia, Eastern Europe, and Latin America, with instances labeled by fluent speakers. |
SkillVerse : Assessing and Enhancing LLMs with Tree Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Language models evolve to tackle complex, multifaceted tasks, requiring granular evaluations . recent studies have focused on leaderboard and benchmark results, but limited interpretability makes it difficult to compare strengths and weaknesses of models. |
| Approach: | They propose an unsupervised tree-structured diagnosis framework for understanding model proficiency in specific abilities with an LLM as a judge. |
| Outcome: | The proposed framework improves model in-context learning and predicts model weaknesses with a 55% success rate compared to the framework without SkillVerse. |
CypherBench: Towards Precise Retrieval over Full-scale Modern Knowledge Graphs in the LLM Era (2025.acl-long)
Copied to clipboard
| Challenge: | Graphs are used for storing open-domain knowledge and domain-specific enterprise data. |
| Approach: | They propose to use property graph views on top of the underlying RDF graph to efficiently query LLMs. |
| Outcome: | The proposed graph views can be efficiently queried by LLMs using Cypher . the proposed graphs have a large schema, overlapping and ambiguous relation types and lack of normalization. |
Empathy Prediction from Diverse Perspectives (2025.acl-long)
Copied to clipboard
Francine Chen, Scott Carter, Tatiana Lau, Nayeli Suseth Bravo, Sumanta Bhattacharyya, Kate Sieck, Charlene C. Wu
| Challenge: | Empathy from perspectives is a prediction tool that uses a rater’s perspective to predict the rater's empathy towards a story. |
| Approach: | They developed a model that uses a rater’s perspective as context for predicting the rater's empathy towards a story. |
| Outcome: | The proposed model improves on the EmpathyFromPerspectives dataset and compares it with baseline models. |
Are LLMs effective psychological assessors? Leveraging adaptive RAG for interpretable mental health screening through psychometric practice (2025.acl-long)
Copied to clipboard
| Challenge: | standardized questionnaires are essential tools for mental health screening, but computational approaches bypass these tools in favor of black-box classification. |
| Approach: | They propose a questionnaire-guided screening framework that bridges psychological practice and computational methods through adaptive Retrieval-Augmented Generation. |
| Outcome: | The proposed framework matches or outperforms state-of-the-art performance on Reddit-based benchmarks and extends to self-harm screening. |
INTERACT: Enabling Interactive, Question-Driven Learning in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) absorb static data without the ability to question and refine knowledge. |
| Approach: | They propose a framework in which a “student” LLM engages a ‘teacher’ LLM through iterative inquiries to acquire knowledge across 1,347 contexts. |
| Outcome: | The proposed framework achieves up to 25% improvement in 1,347 contexts across a wide range of scenarios and LLM architectures, with ‘cold-start’ student models matching static learning baselines in as few as five dialogue turns. |
Circuit Stability Characterizes Language Model Generalization (2025.acl-long)
Copied to clipboard
| Challenge: | Rapid development of state-of-the-art models induce benchmark saturation, while creating more challenging datasets is labor-intensive. |
| Approach: | They propose to introduce circuit stability as a new way to assess model performance. |
| Outcome: | The proposed methods characterize and predict different aspects of generalization. |
Comparing LLM-generated and human-authored news text using formal syntactic theory (2025.acl-long)
Copied to clipboard
| Challenge: | a systematic comparison of LLM-generated and human-authored texts is a topic of growing interest in the field of natural language processing. |
| Approach: | They compare LLM-generated and human-authored New York Times texts using a formal syntactic theory . they use a broad-coverage English resource grammar to analyze the texts . |
| Outcome: | The proposed comparisons reveal systematic differences between human and LLM-generated texts . the authors hope the results will lead to further discoveries about grammatical properties of LLMs . |
Improving Preference Extraction In LLMs By Identifying Latent Knowledge Through Classifying Probes (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are often used as automated judges to evaluate text, but their effectiveness can be hindered by unintentional biases. |
| Approach: | They propose to use linear classifying probes to access LLMs’ latent knowledge and extract more accurate preferences by leveraging differences between contrasting pairs of prompts. |
| Outcome: | The proposed method outperforms generation-based judgements while maintaining similar computational costs. |
White Men Lead, Black Women Help? Benchmarking and Mitigating Language Agency Social Biases in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Social biases manifest in language agency, but there is no comprehensive benchmark for evaluating such biase in language models. |
| Approach: | They propose a benchmark to evaluate language agency biases in large language models . they propose 'Mitigation via Selective Rewrite' to selectively revise parts of generated texts . |
| Outcome: | The proposed language agency bias evaluation benchmark identifies gender, racial, and intersectional biases in 3 recent LLMs. |
AIMSCheck: Leveraging LLMs for AI-Assisted Review of Modern Slavery Statements Across Jurisdictions (2025.acl-long)
Copied to clipboard
Adriana Eufrosina Bora, Akshatha Arodi, Duoyi Zhang, Jordan Bannister, Mirko Bronzi, Arsene Fansi Tchango, Md Abul Bashar, Richi Nayak, Kerrie Mengersen
| Challenge: | Modern slavery acts require corporations to publish annual statements outlining their efforts to combat modern slavery. |
| Approach: | They present annotated datasets from the UK and Canada to enable cross-jurisdictional evaluation. |
| Outcome: | The proposed framework generalizes well across UK and Canadian jurisdictions, demonstrating the potential for broader application in compliance monitoring. |
Collapse of Dense Retrievers: Short, Early, and Literal Biases Outranking Factual Evidence (2025.acl-long)
Copied to clipboard
| Challenge: | Notably, when multiple biases combine, models exhibit catastrophic performance degradation, selecting the answer-containing document in less than 10% of cases over a synthetic biased document without the answer. |
| Approach: | They repurpose a relation extraction dataset to quantify the impact of heuristic biases on retrievers like Dragon+ and Contriever. |
| Outcome: | The proposed models exhibit catastrophic performance degradation when multiple biases combine, selecting the answer-containing document in less than 10% of cases over a synthetic biased document without the answer. |
SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have found that Language Models struggle to fully comprehend and utilize key evidence from the context. |
| Approach: | They propose an inference-time approach that helps LMs focus on key contextual evidence through self-guided explicit highlighting. |
| Outcome: | The proposed method improves on multiple evidence-based QA tasks while maintaining computational efficiency. |
The Male CEO and the Female Assistant: Evaluation and Mitigation of Gender Biases in Text-To-Image Generation of Dual Subjects (2025.acl-long)
Copied to clipboard
| Challenge: | Recent large-scale T2I models like DALLE-3 have made progress in reducing gender stereotypes when generating single-person images. |
| Approach: | They propose a framework that queries T2I models to depict two individuals with gender-stereotyped social identities to evaluate gender biases. |
| Outcome: | The proposed framework reduces gender stereotypes when generating images with more than one person. |
Mitigating Shortcut Learning with InterpoLated Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing shortcut mitigation approaches are model-specific, difficult to tune, computationally expensive, and fail to improve learned representations. |
| Approach: | They propose to interpolate representations of majority examples to include features from intra-class minority examples with shortcut-mitigating patterns. |
| Outcome: | The proposed method improves minority generalization over ERM and state-of-the-art mitigation methods on multiple natural language understanding tasks while preserving accuracy on majority examples. |
Toward Automatic Discovery of a Canine Phonetic Alphabet (2025.acl-long)
Copied to clipboard
| Challenge: | a new algorithm for vocalization communication between dogs is being developed . phonetic units alone are not sufficient to constitute a "language" |
| Approach: | They propose an algorithm that produces a complete alphabet of distinct canine phonemes . the algorithm is expected to function on canines and other animal species . |
| Outcome: | The proposed algorithm produces a complete alphabet of distinct canine phoneme-like units . it is expected to work on canines and other animal species . |
DavIR: Data Selection via Implicit Reward for Large Language Models (2025.acl-long)
Copied to clipboard
Haotian Zhou, Tingkai Liu, Qianli Ma, Yufeng Zhang, Jianbo Yuan, Pengfei Liu, Yang You, Hongxia Yang
| Challenge: | 6% of Alpaca dataset selected with DavIR can steer both LLaMA and Gemma models to produce superior performance compared to the same models trained on the full 52K dataset. |
| Approach: | They propose a model-based data selection method for post-training Large Language Models . they generalize Reducible Holdout Loss to core-set selection problem of causal language modeling . |
| Outcome: | The proposed method can steer both LLaMA and Gemma models to superior performance compared to the same models trained on the full 52K dataset. |
Byte Latent Transformer: Patches Scale Better Than Tokens (2025.acl-long)
Copied to clipboard
Artidoro Pagnoni, Ramakanth Pasunuru, Pedro Rodriguez, John Nguyen, Benjamin Muller, Margaret Li, Chunting Zhou, Lili Yu, Jason E Weston, Luke Zettlemoyer, Gargi Ghosh, Mike Lewis, Ari Holtzman, Srini Iyer
| Challenge: | Existing large language models (LLMs) are trained on bytes, except for tokenization, which groups bytes into a static set of tokens. |
| Approach: | They propose a new byte-level LLM architecture that encodes bytes into dynamically sized patches, which serve as the primary units of computation. |
| Outcome: | The proposed architecture matches tokenization-based models with improvements in inference efficiency and robustness. |
DiffuseDef: Improved Robustness to Adversarial Attacks via Iterative Denoising (2025.acl-long)
Copied to clipboard
| Challenge: | Existing adversarial defense methods for natural language processing still pose challenges to adversarials. |
| Approach: | They propose a novel adversarial defense method that incorporates a diffusion layer as a denoiser between the encoder and the classifier. |
| Outcome: | The proposed method improves over existing adversarial defense methods and achieves state-of-the-art performance against black-box and white-box adversarials. |
Identifying Cellular Niches in Spatial Transcriptomics: An Investigation into the Capabilities of Large Language Models (2025.acl-long)
Copied to clipboard
Huanhuan Wei, Xiao Luo, Hongyi Yu, Jinping Liang, Luning Yang, Lixing Lin, Alexandra Popa, Xiting Yan
| Challenge: | Spatial transcriptomic technologies allow measuring gene expression profile and spatial information of cells in tissues simultaneously. |
| Approach: | They propose a spatial transcriptomic approach to identify spatial niches using a zero-shot large language models by transforming spatial transcriptomics data into spatial context prompts. |
| Outcome: | The proposed model improves performance by leveraging gene expression of neighboring cells/spots, cell type composition, tissue information, and external knowledge. |
Culture Matters in Toxic Language Detection in Persian (2025.acl-long)
Copied to clipboard
| Challenge: | Toxic language detection is crucial for creating safer online environments and limiting the spread of harmful content. |
| Approach: | They compare different methods for toxic language detection in Persian to fine-tune, enrich data, and cross-lingual transfer learning. |
| Outcome: | The language of a country with cultural similarities to Persian yields better results in transfer learning. |
Bitnet.cpp: Efficient Edge Inference for Ternary LLMs (2025.acl-long)
Copied to clipboard
Jinheng Wang, Hansong Zhou, Ting Song, Shijie Cao, Yan Xia, Ting Cao, Jianyu Wei, Shuming Ma, Hongyu Wang, Furu Wei
| Challenge: | 1-bit large language models have spurred interest in ternary LLMs, but efficient edge inference is still scarce. |
| Approach: | They propose an inference system optimized for 1-bit large language models . they propose a new library that facilitates sub-2-bits-per-weight inference . |
| Outcome: | The proposed inference system achieves 6.25x speed increase over full-precision baselines and 2.32x over low-bit baselines. |
Instance-Selection-Inspired Undersampling Strategies for Bias Reduction in Small and Large Language Models for Binary Text Classification (2025.acl-long)
Copied to clipboard
Guilherme Fonseca, Washington Cunha, Gabriel Prenassi, Marcos André Gonçalves, Leonardo Chaves Dutra Da Rocha
| Challenge: | Existing methods to mitigate class imbalanced datasets are limited by existing methods. |
| Approach: | They propose two undersampling methods inspired by state-of-the-art Instance Selection techniques to mitigate class imbalance bias in ATC. |
| Outcome: | The proposed methods reduce classifier bias (56%) across all datasets without effectiveness loss while improving efficiency (1.6x speedup), scalability and reducing carbon emissions (up to 50%). |
Forward Knows Efficient Backward Path: Saliency-Guided Memory-Efficient Fine-tuning of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | a number of fine-tuning approaches are available to improve performance of large language models. |
| Approach: | They propose a memory-efficient method to minimize memory associated with cached intermediate activations. |
| Outcome: | The proposed method minimizes memory associated with cached intermediate activations while preserving accuracy. |
Focus on What Matters: Enhancing Medical Vision-Language Models with Automatic Attention Alignment Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods rely on inference-time interventions, which are limited in attention adaptation or require additional supervision. |
| Approach: | They propose a framework for automatic attention alignment tuning that leverages weak labels from SAM and selectively modifies visually-critical attention heads to improve alignment while minimizing interference. |
| Outcome: | The proposed framework outperforms state-of-the-art models on medical VQA and report generation benchmarks. |
LLMs + Persona-Plug = Personalized LLMs (2025.acl-long)
Copied to clipboard
Jiongnan Liu, Yutao Zhu, Shuting Wang, Xiaochi Wei, Erxue Min, Yu Lu, Shuaiqiang Wang, Dawei Yin, Zhicheng Dou
| Challenge: | Large language models (LLMs) have demonstrated extraordinary capabilities in natural language understanding, generation, and reasoning. |
| Approach: | They propose a plug-and-play LLM model that embeds a user-specific embedding for each individual by modeling her historical contexts through a lightweight plug-in user embedder module. |
| Outcome: | Experiments on various tasks in the language model personalization (LaMP) benchmark show that the proposed model significantly outperforms existing personalized LLM approaches. |
Developmentally-plausible Working Memory Shapes a Critical Period for Language Acquisition (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models possess general linguistic abilities comparable to humans but their efficiency in language acquisition remains far inferior. |
| Approach: | They propose a method that initially constrains working memory during the early stages of training and gradually relaxes this constraint as learning progresses. |
| Outcome: | The proposed method outperforms conventional methods without memory constraints or with static memory constraints. |
IRIS: An Iterative and Integrated Framework for Verifiable Causal Discovery in the Absence of Tabular Data (2025.acl-long)
Copied to clipboard
| Challenge: | Existing statistical methods for causal discovery are expensive, require high-quality structured tabular data, and are often not available for a wide range of NLP applications. |
| Approach: | They propose a framework that combines statistical and large language model methods to discover causal relations from a set of initial variables. |
| Outcome: | The proposed method combines statistical and LLM-based methods to discover known and novel causal relations. |
INJONGO: A Multicultural Intent Detection and Slot-filling Dataset for 16 African Languages (2025.acl-long)
Copied to clipboard
Hao Yu, Jesujoba Oluwadara Alabi, Andiswa Bukula, Jian Yun Zhuang, En-Shiun Annie Lee, Tadesse Kebede Guge, Israel Abebe Azime, Happy Buzaaba, Blessing Kudzaishe Sibanda, Godson Koffi Kalipe, Jonathan Mukiibi, Salomon Kabongo Kabenamualu, Mmasibidi Setaka, Lolwethu Ndolela, Nkiruka Odu, Rooweither Mabuya, Shamsuddeen Hassan Muhammad, Salomey Osei, Sokhar Samb, Dietrich Klakow, David Ifeoluwa Adelani
| Challenge: | Slot-filling and intent detection tasks are well-established tasks in Conversational AI, but current benchmarks for these tasks rely on evaluations of low-resource languages and translations from English benchmarks. |
| Approach: | They propose to use a multilingual, open-source benchmark dataset for 16 African languages with utterances generated by native speakers across diverse domains. |
| Outcome: | The proposed dataset compares multilingual transformer models and prompting large language models (LLMs) with the English language. |
Boosting Long-Context Information Seeking via Query-Guided Activation Refilling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for processing long contexts are ineffective due to their inherent context window limitations and the computational burden of extensive key-value activations. |
| Approach: | They propose a method for processing long context information-seeking tasks via query-guided ACtivation REfilling (ACRE) a bi-layer KV Cache is constructed where the layer-1 cache compactly captures global information and the layer-2 cache provides detailed, localized information. |
| Outcome: | The proposed method achieves significant improvements in both performance and efficiency on a variety of long-context information-seeking datasets. |
Efficient Pretraining Data Selection for Language Models via Multi-Actor Collaboration (2025.acl-long)
Copied to clipboard
Tianyi Bai, Ling Yang, Zhen Hao Wong, Fupeng Sun, Xinlin Zhuang, Jiahui Peng, Chi Zhang, Lijun Wu, Qiu Jiantao, Wentao Zhang, Binhang Yuan, Conghui He
| Challenge: | Efficient data selection is crucial to accelerate the pretraining of language models . limited research has addressed the inherent conflicts between data selection methods . |
| Approach: | They propose a multi-actor collaborative data selection mechanism that prioritizes data based on its specific criterion and updates prioritization rules using the current state of the model. |
| Outcome: | The proposed model accelerates convergence in LM pretraining and achieves an average relative performance gain of 10.5% across multiple language model benchmarks. |
AdaDHP: Fine-Grained Fine-Tuning via Dual Hadamard Product and Adaptive Parameter Selection (2025.acl-long)
Copied to clipboard
| Challenge: | Increasing number of parameters can be challenging under resource-constrained environments. |
| Approach: | They propose a parameter-efficient fine-tuning method with fewer parameters and finer granularity that can adaptively select important parameters for each task. |
| Outcome: | The proposed method can fine-tune important parameters for each task, while maintaining the same weights. |
KG-Agent: An Efficient Autonomous Agent Framework for Complex Reasoning over Knowledge Graph (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to design the interaction strategy between large language models and knowledge graphs (KGs) are not effective for large language model (LLM)s to solve complex tasks due to the large volume and structured format of KG data. |
| Approach: | They propose an LLM-based agent framework that enables small LLMs to actively make decisions over knowledge graphs. |
| Outcome: | The proposed framework outperforms existing methods on in-domain and out-domain datasets using 10K samples. |
Curriculum Debiasing: Toward Robust Parameter-Efficient Fine-Tuning Against Dataset Biases (2025.acl-long)
Copied to clipboard
| Challenge: | Parameter-efficient fine-tuning (PEFT) addresses the memory footprint issue of full fine- tuning by modifying only a subset of model parameters. |
| Approach: | They propose a framework that debiases models in a biased-to-unbiased order and uses only a subset of parameters to modify model parameters. |
| Outcome: | The proposed framework accelerates convergence on unbiased examples by approximately twofold and improves ID and OOD performance by 1.2% and 8.0%, respectively. |
Does Context Matter? ContextualJudgeBench for Evaluating LLM-based Judges in Contextual Settings (2025.acl-long)
Copied to clipboard
| Challenge: | Contextual evaluation is challenging for state-of-the-art judge models . evaluation criteria are often conditional and dependent on practitioner priorities . |
| Approach: | They propose a judge benchmark that evaluates large language models as judges in contexts . they use human annotations and model-based perturbations to build the benchmark . |
| Outcome: | The proposed benchmark aims to evaluate large language models in contexts with 2,000 challenging response pairs. |
On the Reliability of Large Language Models for Causal Discovery (2025.acl-long)
Copied to clipboard
| Challenge: | Existing statistical methods to identify causal relationships from observational data remain elusive. |
| Approach: | They examine the impact of memorization for accurate causal relation prediction, the influence of incorrect causal relations in pre-training data and the contextual nuances that influence LLMs’ understanding of causal relations. |
| Outcome: | The proposed models are effective in recognizing causal relations that occur frequently in pre-training data, but their ability to generalize to new or rare causal relations is limited. |
Value-Spectrum: Quantifying Preferences of Vision-Language Models via Value Decomposition in Social Media Contexts (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Vision-Language Models (VLMs) have broadened the scope of multimodal applications, but evaluations often neglect abstract dimensions such as personality traits and human values. |
| Approach: | They propose a Visual Question Answering (VQA) benchmark based on Schwartz’s value dimensions that capture core human values guiding people’s preferences and actions. |
| Outcome: | The proposed model can be used to evaluate visual question answering (VQA) tasks and to simulate diverse personas. |
TeRDy: Temporal Relation Dynamics through Frequency Decomposition for Temporal Knowledge Graph Completion (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for temporal knowledge graph completion struggle to capture long-term changes and short-term variability of relations. |
| Approach: | They propose a method that captures temporal relational dynamics by time-invariant embeddings and time-outvariant time-variant embeddedding. |
| Outcome: | The proposed method outperforms state-of-the-art methods on benchmark datasets. |
Incorporating Domain Knowledge into Materials Tokenization (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in language models have expanded their applications in materials science, but they often produce excessive fragmentation and semantic loss. |
| Approach: | They propose a frequency-centric tokenization approach that integrates material knowledge into tokenization. |
| Outcome: | The proposed tokenization approach outperforms existing tokenization methods and achieves an average performance gain of 4% and 2% in the generation and classification tasks. |
PIG: Privacy Jailbreak Attack on LLMs via Gradient-based Iterative In-Context Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to evaluate privacy leakage in LLMs use memorized prefixes or simple instructions to extract data, which well-aligned models can easily block. |
| Approach: | They propose a framework targeting Personally Identifiable Information (PII) that uses in-context learning to build a privacy context and iteratively updates it with three gradient-based strategies to elicit target PII. |
| Outcome: | The proposed framework outperforms baseline methods and achieves state-of-the-art (SoTA) results on four white-box and two black-box LLMs. |
Agents Under Siege: Breaking Pragmatic Multi-Agent LLM Systems with Optimized Prompt Attacks (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-agent LLMs are prone to adversarial attacks because of constraints such as limited token bandwidth and latency between message delivery. |
| Approach: | They propose a permutation-invariant adversarial attack that optimizes prompt distribution across latency and bandwidth constraints to bypass distributed safety mechanisms within the system. |
| Outcome: | The proposed method outperforms conventional attacks by up to 7 on multiple models. |
Semantic-Eval : A Semantic Comprehension Evaluation Framework for Large Language Models Generation without Training (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have emerged as key drivers of progress in the field of natural language processing. |
| Approach: | They propose a framework that assesses LLM-generated text based on semantic understanding. |
| Outcome: | The proposed framework surpasses traditional evaluation metrics and lags behind GPT-4. |
Between Circuits and Chomsky: Pre-pretraining on Formal Languages Imparts Linguistic Biases (2025.acl-long)
Copied to clipboard
| Challenge: | Pretraining language models on formal language can improve their acquisition of natural language, but it is computationally limited. |
| Approach: | They hypothesize that effective transfer occurs when the formal language captures hierarchical dependencies present in natural language and remains within the computational limitations of the model architecture. |
| Outcome: | The proposed model can achieve lower loss on natural language and better generalization with a 33% smaller token budget than training on a matched amount of natural language. |
When to Speak, When to Abstain: Contrastive Decoding with Abstention (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) demonstrate exceptional performance across diverse tasks by leveraging pre-trained (parametric) and external (contextual) knowledge. |
| Approach: | They propose a training-free decoding method that allows LLMs to generate responses when relevant knowledge is available and to abstain otherwise. |
| Outcome: | The proposed method can generate responses when relevant knowledge is available and abstain otherwise. |
On the Risk of Evidence Pollution for Malicious Social Text Detection in the Era of LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Evidence-enhanced detectors are able to detect malicious social text, but they are prone to evidence pollution. |
| Approach: | They propose three defense strategies to mitigate evidence pollution by large language models by machine-generated text detection and a mixture of experts. |
| Outcome: | The proposed defense strategies could mitigate evidence pollution, but they faced limitations for practical employment. |
Investigating and Extending Homans’ Social Exchange Theory with Large Language Model based Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Social exchange theory (SET) is widely recognized as a basic framework for understanding human interactions and interactions. |
| Approach: | They propose to use large language models to study Homans’ social exchange theory (SET) by constructing a virtual society composed of three LLM agents and having them engage in a social exchange game to observe their behaviors. |
| Outcome: | The proposed model extends Homans’ SET with LLM-based agents and demonstrates consistency between the agent and human behavior. |
A Drop-In Solution for On-the-Fly Adaptation of Speculative Decoding in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are highly memory-intensive when performing real-time inference. |
| Approach: | They propose a technique that allows for speculative decoding to be run on the fly to maximize the efficiency of LLM inferences. |
| Outcome: | The proposed solution can lead to 3.55-16.48% speed improvement over the standard speculative decoding, and 1.2-3.4 over the default LLMs. |
If Attention Serves as a Cognitive Model of Human Memory Retrieval, What is the Plausible Memory Representation? (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work in computational psycholinguistics has revealed intriguing parallels between attention mechanisms and human memory retrieval, focusing primarily on vanilla Transformers that operate on token-level representations. |
| Approach: | They propose that the attention mechanism of Transformer Grammar (TG) can serve as a cognitive model of human memory retrieval using Normalized Attention Entropy (NAE) they propose that TG's attention can implement a human memory-retrieval theory known as cue-based retrieval . |
| Outcome: | The attention mechanism of Transformer Grammar (TG) achieves superior predictive power for self-paced reading times compared to vanilla Transformer’s, with further analyses revealing independent contributions from both models. |
Aligning VLM Assistants with Personalized Situated Cognition (2025.acl-long)
Copied to clipboard
Yongqi Li, Shen Zhou, Xiaohu Li, Xin Miao, Jintao Wen, Mayi Xu, Jianhao Chen, Birong Pan, Hankun Kang, Yuanyuan Zhu, Ming Zhong, Tieyun Qian
| Challenge: | Existing studies on vision-language models aligned with general human objectives have not been successful because people with diversified backgrounds have different cognition even in the same situation. |
| Approach: | They propose to characterize individuals based on the sociological concept of Role-Set and then evaluate their actions to see whether personalized alignment is achieved. |
| Outcome: | The proposed framework constructs a cognition-aware and action-based reward model for personalized alignment. |
Attention Entropy is a Key Factor: An Analysis of Parallel Context Encoding with Full-attention-based Pre-trained Language Models (2025.acl-long)
Copied to clipboard
Zhisong Zhang, Yan Wang, Xinting Huang, Tianqing Fang, Hongming Zhang, Chenlong Deng, Shuaiyi Li, Dong Yu
| Challenge: | Large language models have demonstrated remarkable performance across a wide range of language tasks due to their remarkable ability in context modeling. |
| Approach: | They propose to use parallel context encoding to reduce attention entropy by incorporating attention sinks and selective mechanisms to reduce irregular attention . they also propose to incorporate attention sink mechanisms into the parallel encoded context to reduce the irregular attention. |
| Outcome: | The proposed methods lower irregular attention entropy and narrow performance gaps. |
Faster Speculative Decoding via Effective Draft Decoder with Pruned Candidate Tree (2025.acl-long)
Copied to clipboard
| Challenge: | Effective Draft Decoder (EDD) is a powerful decoding method that generates more accurate draft tokens by leveraging the encoding results as soft prompts. |
| Approach: | They propose an effective draft decoder which treats the LLM as a powerful encoder and generates more accurate draft tokens by leveraging the encoding results as soft prompts. |
| Outcome: | The proposed method significantly improves the performance of large language models and reduces inference latency. |
Selecting and Merging: Towards Adaptable and Scalable Named Entity Recognition with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to align large language models with information extraction tasks are costly and not all training data benefits target domains. |
| Approach: | They propose a framework which dynamically Selects and Merges expert models at inference time and combines experts beneficial to target domains. |
| Outcome: | The proposed framework outperforms the unified model by 10% on multiple benchmarks. |
Embracing Imperfection: Simulating Students with Diverse Cognitive Levels Using LLM-based Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are becoming increasingly popular in education, enabling researchers to simulate students' learning patterns and learning patterns. |
| Approach: | They propose a training-free framework for student simulation that takes into account student cognitive diversity and realism. |
| Outcome: | The proposed model outperforms baseline models and achieves 100% improvement in simulation accuracy and realism. |
CADReview: Automatically Reviewing CAD Programs with Error Detection and Correction (2025.acl-long)
Copied to clipboard
| Challenge: | Computer-aided design (CAD) is crucial in prototyping 3D objects through geometric instructions. |
| Approach: | They propose a CAD review task to automatically detect and correct potential errors . they propose CAD program repairer framework to provide helpful feedback on error correction . |
| Outcome: | The proposed framework outperforms existing MLLMs in detecting errors and providing feedback on error correction. |
Think&Cite: Improving Attributed Text Generation with Self-Guided Tree Search and Progress Reward Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are prone to hallucinations and producing factually incorrect information. |
| Approach: | They propose a framework that allows LLMs to generate citations that provide evidence for any statement. |
| Outcome: | The proposed framework outperforms baseline approaches on three datasets and significantly outperformed baseline approaches. |
The Lawyer That Never Thinks: Consistency and Fairness as Keys to Reliable AI (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly used in high-stakes domains like law and research. |
| Approach: | They evaluate six leading Large Language Models on rationality, stability, and ethical fairness through reasoning tests, legal challenges, and bias-sensitive scenarios. |
| Outcome: | The models perform well on reasoning tests, legal challenges, and bias-sensitive scenarios. |
Polishing Every Facet of the GEM: Testing Linguistic Competence of LLMs and Humans in Korean (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have focused on linguistic competence of language models with grammatical knowledge. |
| Approach: | They propose to use grammar as a measurable proxy to assess linguistic competence of large language models (LLMs) . |
| Outcome: | The proposed model aims to assess the linguistic competence of large language models (LLMs) and humans in Korean. |
SpeechFake: A Large-Scale Multilingual Speech Deepfake Dataset Incorporating Cutting-Edge Generation Methods (2025.acl-long)
Copied to clipboard
| Challenge: | Existing speech deepfake datasets are limited in scale and diversity, making it challenging to train models that can generalize well to unseen deepfakkes. |
| Approach: | They propose a large-scale speech deepfake dataset that includes over 3 million deepfak samples, totaling more than 3,000 hours of audio, generated using 40 different speech synthesis tools. |
| Outcome: | The proposed dataset includes over 3 million deepfake samples, totaling more than 3,000 hours of audio, generated using 40 different speech synthesis tools. |
ReflectionCoder: Learning from Reflection Sequence for Enhanced One-off Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to enhance code generation performance include integrating compiler feedback. |
| Approach: | They propose a method that integrates compiler feedback to improve one-off code generation performance. |
| Outcome: | The proposed method improves one-off code generation performance on three benchmarks and can be applied to other domains that focus on final results and require long reasoning paths. |
InvestAlign: Overcoming Data Scarcity in Aligning Large Language Models with Investor Decision-Making Processes Under Herd Behavior (2025.acl-long)
Copied to clipboard
| Challenge: | relying on authentic data for Supervised Fine-Tuning (SFT) is costly and expensive. |
| Approach: | They propose a framework that constructs high-quality SFT datasets by leveraging theoretical solutions to similar and simple optimal investment problems rather than the complex scenarios. |
| Outcome: | The proposed framework achieves faster parameter convergence than using real-user data, suggesting superior learning efficiency. |
Enhancing Neural Machine Translation Through Target Language Data: A kNN-LM Approach for Domain Adaptation (2025.acl-long)
Copied to clipboard
Abudurexiti Reheman, Hongyu Liu, Junhao Ruan, Abudukeyumu Abudula, Yingfeng Luo, Tong Xiao, JingBo Zhu
| Challenge: | Neural machine translation (NMT) has made significant progress in recent years, yet often suffers from translating in new domains, which is called domain adaptation. |
| Approach: | They propose a method that leverages semantically similar target language sentences in the kNN framework and generates a probability distribution over these sentences during decoding. |
| Outcome: | The proposed method generates a probability distribution over similar target language sentences and then interpolates with the model’s distribution. |
Multi-level Relevance Document Identifier Learning for Generative Retrieval (2025.acl-long)
Copied to clipboard
Fuwei Zhang, Xiaoyu Liu, Xinyu Jia, Yingfei Zhang, Shuai Zhang, Xiang Li, Fuzhen Zhuang, Wei Lin, Zhao Zhang
| Challenge: | Existing methods generate DocIDs based on textual content, which may result in weak semantic connections for similar documents due to variations in expression. |
| Approach: | They propose a new retrieval paradigm that generates unique document identifiers . they propose to use queries as a bridge to connect documents with varying relevance levels . |
| Outcome: | The proposed approach outperforms existing methods on multilingual e-commerce search datasets. |
EfficientQAT: Efficient Quantization-Aware Training for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Quantization-aware training (QAT) is a low-bit training solution that requires substantial training resources. |
| Approach: | They propose an algorithm that reduces memory consumption by low-bit representations with minimal accuracy loss. |
| Outcome: | EfficientQAT achieves 2-bit Llama-2-70B model on single GPU in 41 hours . compared to previous methods, it obtains model with less than 3 points accuracy degradation . |
Exploring How Generative MLLMs Perceive More Than CLIP with the Same Vision Encoder (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies show that CLIP models struggle with visual reasoning tasks . despite the success of Contrastive Language-Image Pretraining, there are still limitations . |
| Approach: | They propose to use a visual encoder to train CLIP-like models for fine-grained visual reasoning tasks. |
| Outcome: | The proposed models outperform CLIP-like encoders in visual reasoning tasks . the study highlights the importance of VLM architectural choices . |
NexusSum: Hierarchical LLM Agents for Long-Form Narrative Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | Summarizing long-form narratives requires capturing intricate plotlines, character interactions, and thematic coherence over tens of thousands of tokens. |
| Approach: | They propose a multi-agent LLM framework for narrative summarization that processes long-form text through a structured pipeline without fine-tuning. |
| Outcome: | The proposed framework achieves 30.0% improvement in BERTScore across books, movies, and TV scripts. |
HAIC: Improving Human Action Understanding and Generation with Better Captions for Multi-modal Large Language Models (2025.acl-long)
Copied to clipboard
Xiao Wang, Jingyun Hua, Weihong Lin, Yuanxing Zhang, Fuzheng Zhang, Jianlong Wu, Di Zhang, Liqiang Nie
| Challenge: | Existing studies have shown that high-quality video captions can improve MLLMs' performance on videos involving human actions. |
| Approach: | They propose a data annotation pipeline to collect videos featuring clear human actions from the Internet and annotate them in a standardized caption format that uses human attributes to distinguish individuals. |
| Outcome: | The proposed pipeline combines two datasets to evaluate human action understanding. |
Uni-Retrieval: A Multi-Style Retrieval Framework for STEM’s Education (2025.acl-long)
Copied to clipboard
| Challenge: | Current retrieval models focus on natural text-image retrieval, which is insufficient for STEM education contexts due to ambiguities in the retrieval process. |
| Approach: | They propose a diverse expression retrieval task tailored to educational scenarios . they extract query style features as prototypes and build a continuously updated Prompt Bank . |
| Outcome: | The proposed model outperforms existing retrieval models in most retrieval tasks. |
DenseLoRA: Dense Low-Rank Adaptation of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Low-rank adaptation (LoRA) is an efficient approach for adapting large language models (LLMs) but many of the weights in these matrices are redundant, leading to inefficiencies in parameter utilization. |
| Approach: | They propose a low-rank adaptation approach that fine-tunes two low-ranked matrices and adapts them through a dense low-Rank matrix, improving parameter utilization and adaptation efficiency. |
| Outcome: | The proposed approach achieves 83.8% accuracy with only 0.01% of trainable parameters compared to LoRA's 80.8% with 0.70% of trainability parameters on LLaMA3-8B. |
Exploring the Potential of LLMs as Personalized Assistants: Dataset, Evaluation, and Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Personalized AI assistants are a challenging application that intertwines multiple problems in LLM research. |
| Approach: | They propose a Llama-3.2-based automated evaluation model that matches human preferences to a conversational dataset. |
| Outcome: | HiCUPID provides a conversational dataset tailored for personalization . the evaluation model closely mirrors human preferences, the researchers show . |
Cracking Factual Knowledge: A Comprehensive Analysis of Degenerate Knowledge Neurons in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are believed to store extensive factual knowledge, yet the mechanisms of knowledge storage in LLMs remain largely unexplored. |
| Approach: | They propose that some multi-layer perceptron neurons can store "knowledge". |
| Outcome: | The proposed model can store "knowledge" in multi-layer perceptron neurons, but not redundancy. |
Towards Context-Robust LLMs: A Gated Representation Fine-tuning Approach (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) enhanced with external contexts face challenges in handling imperfect evidence. |
| Approach: | They propose a framework that can balance internal knowledge with external contexts . they propose gating mechanisms and low-rank representation adapters to adjust hidden representations based on a lightweight intervention function . |
| Outcome: | The proposed model can effectively balance internal knowledge with external context, similar to human cognitive processes. |
On Support Samples of Next Word Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Language models excel in various tasks by making complex decisions, but understanding the rationale behind these decisions remains a challenge. |
| Approach: | They investigate data-centric interpretability in language models by focusing on the next-word prediction task. |
| Outcome: | The proposed model supports or deteres specific predictions, while non-support samples play a critical role in generalization and representation learning. |
WebWalker: Benchmarking LLMs in Web Traversal (2025.acl-long)
Copied to clipboard
Jialong Wu, Wenbiao Yin, Yong Jiang, Zhenglin Wang, Zekun Xi, Runnan Fang, Linhai Zhang, Yulan He, Deyu Zhou, Pengjun Xie, Fei Huang
| Challenge: | Large Language Models (LLMs) have demonstrated impressive capabilities across a wide range of natural language processing tasks. |
| Approach: | They propose a benchmark to assess the ability of LLMs to perform web traversal by using an explore-critic paradigm. |
| Outcome: | The proposed framework mimics human-like web navigation through an explore-critic paradigm and demonstrates the effectiveness of RAG combined with WebWalker in real-world scenarios. |
From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing watermarking methods face limitations that hinder their effectiveness in diverse and adversarial scenarios. |
| Approach: | They propose a symbiotic watermarking framework with three strategies: serial, parallel, and hybrid. |
| Outcome: | The proposed framework outperforms baselines and achieves state-of-the-art (SOTA) performance. |
AutoGUI: Scaling GUI Grounding with Automatic Functionality Annotations from LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing datasets for UI-VLMs contain large-scale context-free element annotations or contextualized functional descriptions for elements at a small scale. |
| Approach: | They propose an auto-annotation pipeline that generates massive UI element functionality annotations based on UI content changes induced by interacting with the elements. |
| Outcome: | The proposed pipeline improves accuracy and scales well with human evaluation of a high-quality AutoGUI-704k dataset. |
Introducing Graph Context into Language Models through Parameter-Efficient Fine-Tuning for Lexical Relation Mining (2025.acl-long)
Copied to clipboard
| Challenge: | Pre-trained language models can effectively mine lexical relations between word pairs . however, graph features and semantic knowledge of pre-tried models are lacking in the task. |
| Approach: | They propose a parameter-efficient fine-tuning method which integrates graph features and semantic representations for lexical relation classification and lexic entailment tasks. |
| Outcome: | The proposed method integrates graph features and semantic representations for lexical relation mining tasks. |
S-RAG: A Novel Audit Framework for Detecting Unauthorized Use of Personal Data in RAG Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Retrieval-Augmented Generation (RAG) systems rely on external data for accurate and context-specific responses. |
| Approach: | They propose a framework that enables users to determine whether their textual data has been utilized in RAG systems even in black-box settings with no prior system knowledge. |
| Outcome: | The proposed framework achieves an improvement in Accuracy by 19.9% while maintaining strong performance under adversarial defenses. |
Praetor: A Fine-Grained Generative LLM Evaluator with Instance-Level Customizable Evaluation Criteria (2025.acl-long)
Copied to clipboard
Yongqi Leng, Renren Jin, Yue Chen, Zhuowen Han, Ling Shi, Jianxiang Peng, Lei Yang, Juesi Xiao, Deyi Xiong
| Challenge: | Existing evaluation methods are inadequate to evaluate large language models (LLMs). |
| Approach: | They propose a fine-grained generative LLM evaluator with instance-level customazable evaluation criteria that can be used to evaluate large language models. |
| Outcome: | The proposed model outperforms existing LLM evaluators and instruction-tuned LLMs on multiple benchmarks and sets new SOTA results. |
Mitigating Confounding in Speech-Based Dementia Detection through Weight Masking (2025.acl-long)
Copied to clipboard
| Challenge: | Pre-trained neural language models fine-tuned on AD transcripts perform well, but little research has explored the effects of the gender of the speakers represented by these transcripts. |
| Approach: | They propose to use the Extended Confounding Filter and the Dual Filter to isolate and ablate weights associated with gender in dementia datasets. |
| Outcome: | The proposed methods overfit to training data distributions and disrupt gender-related weights, with the trade-off of slightly reduced dementia detection performance. |
MCS-Bench: A Comprehensive Benchmark for Evaluating Multimodal Large Language Models in Chinese Classical Studies (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal Large Language Models (MLLMs) have advanced visual and language understanding, but their potential in Chinese Classical Studies (CCS) remains underexplored due to the lack of specialized benchmarks. |
| Approach: | They propose to develop a multimodal benchmark specifically designed for Chinese Classical Studies across multiple subdomains to bridge this gap. |
| Outcome: | The proposed benchmark spans seven core subdomains with a total of 45 meticulously designed tasks. |
The Knowledge Microscope: Features as Better Analytical Lenses than Neurons (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have shown that features are superior analytical units for understanding factual knowledge in Language Models. |
| Approach: | They propose a feature-based editing method that decomposes neurons into features rather than neurons to understand the mechanisms of factual knowledge in Language Models. |
| Outcome: | The proposed method demonstrates superior performance over neuron-based approaches in erasing privacy-sensitive information from LMs. |
From Real to Synthetic: Synthesizing Millions of Diversified and Complicated User Instructions with Attributed Grounding (2025.acl-long)
Copied to clipboard
| Challenge: | a pursuit of diverse, complex, and large-scale instruction data is crucial for automatically aligning large language models . authors: methods that generate synthetic instructions at scale suffer from limited grounding sources . attributed grounding is a technique that can be used to align language models with human . |
| Approach: | They synthesize 1 million instructions using attributed grounding and a bottom-up synthesis process that leverages web documents to generate a situation, then a meaningful instruction. |
| Outcome: | The proposed framework achieves leading performance on benchmarks and scales with more web corpora. |
PrivaCI-Bench: Evaluating Privacy with Contextual Integrity and Legal Compliance (2025.acl-long)
Copied to clipboard
Haoran Li, Wenbin Hu, Huihao Jing, Yulin Chen, Qi Hu, Sirui Han, Tianshu Chu, Peizhao Hu, Yangqiu Song
| Challenge: | Recent advances in generative large language models (LLMs) have enabled wider applicability, accessibility, and flexibility. |
| Approach: | They propose a contextual privacy evaluation benchmark that covers the entire relevant social context through private information flows. |
| Outcome: | The proposed benchmarks cover legal compliance, real court cases, privacy policies, and synthetic data. |
Unveiling Environmental Impacts of Large Language Model Serving: A Functional Unit View (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies benchmark carbon emissions but lack a standardized basis for comparison across different model configurations. |
| Approach: | They propose a functional unit (FU) as a standardized basis for comparison across different model configurations and introduce FUEL, the first FU-based framework for evaluating LLM serving’s environmental impact. |
| Outcome: | The proposed framework uncovers key insights and trade-offs in reducing carbon emissions by optimizing model size, quantization strategy, and hardware choice, paving the way for more sustainable LLM serving. |
ExpeTrans: LLMs Are Experiential Transfer Learners (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies provide large language models with textual task-solving experiences via prompts to improve their performance. |
| Approach: | They propose to use prompts to provide LLMs with textual task-solving experiences during their inference stage. |
| Outcome: | The proposed framework improves the performance of large language models on 13 datasets. |
Cool-Fusion: Fuse Large Language Models without Training (2025.acl-long)
Copied to clipboard
| Challenge: | Cool-Fusion is a simple yet effective approach to combine two or more heterogeneous large language models . |
| Approach: | They propose a method that fuses the knowledge of two or more heterogeneous large language models to leverage complementary strengths. |
| Outcome: | The proposed method increases accuracy from three strong source LLMs on GSM8K by 17.4%. |
DAPE V2: Process Attention Score as Feature Map for Length Extrapolation (2025.acl-long)
Copied to clipboard
Chuanyang Zheng, Yihang Gao, Han Shi, Jing Xiong, Jiankai Sun, Jingyao Li, Minbin Huang, Xiaozhe Ren, Michael Ng, Xin Jiang, Zhenguo Li, Yu Li
| Challenge: | Extensive experiments demonstrate that treating attention as a feature map and applying convolution as . a processing method significantly enhances Transformer performance. |
| Approach: | They propose to use the convolution operator to mimic the processing methods in computer vision to treat attention as a feature map and apply it to neighboring attention scores across different heads. |
| Outcome: | The proposed model can be adapted to various attention-related models and achieves high performance. |
MuSC: Improving Complex Instruction Following with Multi-granularity Self-Contrastive Training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for complex instruction-following with elaborate constraints rely on a weaker model, especially GPT-4, limiting their application. |
| Approach: | They propose a Multi-granularity Self-Contrastive Training framework to improve instruction alignment without relying on a stronger model. |
| Outcome: | The proposed framework improves instruction-following with elaborate constraints without external supervision on coarse and fine granularity. |
LongReD: Mitigating Short-Text Degradation of Long-Context Large Language Models via Restoration Distillation (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have extended context windows through scaling positional encodings and lightweight continual pre-training, but performance degradation is still not fully explored. |
| Approach: | They propose a novel approach to reduce short-text performance degradation by minimizing distribution drift in hidden states and attention scores. |
| Outcome: | The proposed approach minimizes the distribution discrepancy between the extended and original models while maintaining or even enhancing the model's long-context abilities. |
APB: Accelerating Distributed Long-Context Inference by Passing Compressed Context Blocks across GPUs (2025.acl-long)
Copied to clipboard
Yuxiang Huang, Mingye Li, Xu Han, Chaojun Xiao, Weilin Zhao, Sun Ao, Hao Zhou, Jie Zhou, Zhiyuan Liu, Maosong Sun
| Challenge: | Long-context inference is crucial for advancing large language models, but its prefill speed remains a bottleneck. |
| Approach: | They propose an efficient long-context inference framework that leverages multi-host approximate attention to enhance prefill speed. |
| Outcome: | The proposed framework achieves speedups of 9.2, 4.2, and 1.6 without any degradation in performance. |
PPT: A Minor Language News Recommendation Model via Cross-Lingual Preference Pattern Transfer (2025.acl-long)
Copied to clipboard
| Challenge: | Existing systems for minor language news recommendation lack interaction with content . linguistic gap can lead to inaccurate modeling of minor language content despite strong English capability . |
| Approach: | They propose a minor language news recommendation model by cross-lingual preference pattern transfer . their model employs the widely used two-tower architecture and large language model as the backbone of the news encoder . |
| Outcome: | The proposed model outperforms existing models on 15 minor languages. |
GainRAG: Preference Alignment in Retrieval-Augmented Generation through Gain Signal Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to retrieve information from large language models (LLMs) but they fail to address the preference gap between retrievers and LLMs. |
| Approach: | They propose a retrieval module that dynamically injects retrieved information into the input context of large language models (LLMs) This approach aligns the retriever’s and LLM’s preferences by defining a new metric, “gain”, which measure how well an input passage contributes to correct outputs. |
| Outcome: | The proposed approach has shown significant success in various NLP tasks, but there is a preference gap between retrievers and LLMs. |
Top-n𝜎: Eliminating Noise in Logit Space for Robust Token Sampling of LLM (2025.acl-long)
Copied to clipboard
| Challenge: | Existing sampling methods that are sensitive to temperature scaling fail to distinguish between diversity and noise. |
| Approach: | They propose a method that identifies informative tokens by eliminating noise directly in logit space and a new sampling method that is temperature-invariant. |
| Outcome: | The proposed method outperforms existing methods with significant improvements in reasoning and creative writing tasks. |
SCOPE: Optimizing Key-Value Cache Compression in Long-context Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Excessive compression during the prefill phase impairs comprehension of reasoning tasks . SCOPE is a framework that performs KV cache optimization during the decoding and prefill phases . |
| Approach: | They propose a framework that performs optimization during the prefill and decoding phases . they propose enabling a sliding strategy to select essential heavy hitters for the decoding phase . |
| Outcome: | Experiments show that SCOPE can optimize key-value cache for long-context generation tasks . the framework can preserve essential information while minimizing memory usage and transfer . |
Mitigating Non-Representative Prototypes and Representation Bias in Few-Shot Continual Relation Extraction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for few-shot continual relation extraction (FCRE) face two main challenges: non-representative prototypes and representation bias. |
| Approach: | They propose to use General Orthogonal Frame to create robust class prototypes . they also utilize label description representations as global class representatives . |
| Outcome: | The proposed method outperforms state-of-the-art methods on well-known benchmarks on well known FCRE benchmarks. |
MoQAE: Mixed-Precision Quantization for Long-Context LLM Inference via Mixture of Quantization-Aware Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to optimize large language models for long-context inference are inefficient and consume memory. |
| Approach: | They propose a mixed-precision quantization method via mixture of experts that inputs tokens into router chunk by chunk to reduce inference overhead. |
| Outcome: | The proposed method outperforms state-of-the-art KV cache quantization methods on multiple benchmark datasets. |
PrivacyRestore: Privacy-Preserving Inference in Large Language Models via Privacy Removal and Restoration (2025.acl-long)
Copied to clipboard
| Challenge: | Existing privacy protection methods for large language models suffer from performance degradation or large inference time overhead. |
| Approach: | They propose a plug-and-play method to protect the privacy of user inputs during LLM inference . they use offline restoration vectors to train restoration vector for each privacy span type . |
| Outcome: | The proposed method can prevent the linear growth of the privacy budget. |
Meta-rater: A Multi-dimensional Data Selection Method for Pre-training Language Models (2025.acl-long)
Copied to clipboard
Xinlin Zhuang, Jiahui Peng, Ren Ma, Yinfan Wang, Tianyi Bai, Xingjian Wei, Qiu Jiantao, Chi Zhang, Ying Qian, Conghui He
| Challenge: | composition of pre-training datasets for large language models remains undisclosed . current methods for evaluating data quality are limited by single-dimensional evaluation or redundancy-focused strategies. |
| Approach: | They propose a multi-dimensional data selection method that integrates dimensions with existing quality metrics through learned optimal weightings. |
| Outcome: | The proposed method doubles convergence speed for 1.3B model models and improves downstream task performance by 3.23%. |
GuessArena: Guess Who I Am? A Self-Adaptive Framework for Evaluating LLMs in Domain-Specific Knowledge and Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods for large language models rely on static benchmarks and standardized evaluation protocols. |
| Approach: | They propose an adaptive evaluation framework that integrates dynamic domain knowledge modeling with progressive reasoning assessment to improve evaluation fidelity. |
| Outcome: | Empirical results show that the framework distinguishes LLMs in terms of domain knowledge coverage and reasoning chain completeness. |
Sample-Efficient Human Evaluation of Large Language Models via Maximum Discrepancy Competition (2025.acl-long)
Copied to clipboard
Kehua Feng, Keyan Ding, Tan Hongzhi, Kede Ma, Zhihua Wang, Shuangquan Guo, Cheng Yuzhou, Ge Sun, Guozhou Zheng, Qiang Zhang, Huajun Chen
| Challenge: | Existing methods for evaluation of large language models are inefficient and inefficient due to inaccuracy of standard metrics in human perception of text quality and inefficiency in sampling informative test examples. |
| Approach: | They propose a sample-efficient human evaluation method for large language models based on the principle of MAximum Discrepancy (MAD) competition. |
| Outcome: | The proposed method achieves the “golden” ranking of LLMs with a minimum set of input instructions, which in turn reveal their relative strengths and weaknesses. |
DTCRS: Dynamic Tree Construction for Recursive Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | Recursive summarization (RAG) is an important method for mitigating large model hallucinations and enhancing answer interpretability. |
| Approach: | They propose a method that dynamically generates summary trees based on document structure and query semantics. |
| Outcome: | The proposed method significantly reduces summary tree construction time and achieves substantial improvements across three QA tasks. |
A Generative Adaptive Replay Continual Learning Model for Temporal Knowledge Graph Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Continual Learning (CL)-based Temporal Knowledge Graph Reasoning methods are incomplete and reorganize historical facts without preserving historical knowledge. |
| Approach: | They propose a method which generates and adaptively replays historical entity distributions from the whole historical context. |
| Outcome: | The proposed method outperforms baselines in reasoning and mitigating forgetting. |
ARise: Towards Knowledge-Augmented Reasoning via Risk-Adaptive Search (2025.acl-long)
Copied to clipboard
Yize Zhang, Tianshu Wang, Sirui Chen, Kun Wang, Xingyu Zeng, Hongyu Lin, Xianpei Han, Le Sun, Chaochao Lu
| Challenge: | Large language models (LLMs) have impressive capabilities but their application in open-ended, knowledge-intensive, complex reasoning scenarios is limited. |
| Approach: | They propose a framework that integrates risk assessment of intermediate reasoning states with dynamic retrieval-augmented generation within a Monte Carlo tree search paradigm. |
| Outcome: | The proposed framework outperforms the state-of-the-art KAR methods by up to 23.10% and the latest RAG-equipped large reasoning models by upto 25.37%. |
PKAG-DDI: Pairwise Knowledge-Augmented Language Model for Drug-Drug Interaction Event Text Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Drug-drug interactions arise when multiple drugs are administered concurrently. |
| Approach: | They propose a pairwise knowledge-augmented generative method for DDIE text generation that integrates biological functions from a knowledge set into a language model. |
| Outcome: | The proposed method outperforms existing methods in DDIE text generation on two professional datasets. |
Knowledge-Augmented Multimodal Clinical Rationale Generation for Disease Diagnosis with Small Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models struggle to balance predictive accuracy with human-understandable rationales. |
| Approach: | They propose to enhance LLMs by leveraging rationale distillation and domain knowledge injection for trustworthy multimodal rationale generation. |
| Outcome: | Experiments on real-world medical datasets show that ClinRaGen achieves state-of-the-art performance in disease diagnosis and rationale generation. |
TWIST: Text-encoder Weight-editing for Inserting Secret Trojans in Text-to-Image Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Trojan attacks require extensive training data and poor generalization, limiting effectiveness and scalability. |
| Approach: | They propose a method for embedding Trojans into plugins using a single edit layer . they find that the method reduces modified parameters by 8-fold and cuts injection time to 25 seconds . |
| Outcome: | The proposed method achieves an average attack success rate of 91%, a 78% improvement over the state-of-the-art (SOTA) method. |
Frictional Agent Alignment Framework: Slow Down and Don’t Break Things (2025.acl-long)
Copied to clipboard
| Challenge: | Common preference alignment methods excel in static settings, but struggle in dynamic collaborative tasks where explicit signals of interlocutor beliefs are sparse and skewed. |
| Approach: | They propose a Frictional Agent Alignment Framework to generate precise, context-aware friction that prompts deliberation and re-examination of existing evidence. |
| Outcome: | The proposed framework outperforms existing methods in producing concise, interpretable friction and in OOD generalization. |
Powerformer: Efficient and High-Accuracy Privacy-Preserving Language Model with Homomorphic Encryption (2025.acl-long)
Copied to clipboard
| Challenge: | a new privacy-preserving language model, Powerformer, is designed to reduce computation overhead while maintaining model performance. |
| Approach: | They propose an efficient homomorphic encryption-based privacy-preserving language model . it incorporates three key techniques to optimize encrypted computations . |
| Outcome: | The proposed model achieves 45% reduction in computation time compared to state-of-the-art models . authors say the model preserves data privacy and AI capabilities in MLaaS environments . |
Beware of Your Po! Measuring and Mitigating AI Safety Risks in Role-Play Fine-Tuning of LLMs (2025.acl-long)
Copied to clipboard
Weixiang Zhao, Yulin Hu, Yang Deng, Jiahe Guo, Xingyu Sui, Xinyang Han, An Zhang, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu
| Challenge: | Existing role-play fine-tuning techniques improve role adaptability but may degrade safety performance, especially for villainous characters. |
| Approach: | They propose safety-aware Role-Play Fine-Tuning (SaRFT) to balance role-playing capabilities and safety. |
| Outcome: | The proposed method outperforms state-of-the-art baselines under both LoRA and full-parameter fine-tuning settings. |
Can Graph Neural Networks Learn Language with Extremely Weak Text Supervision? (2025.acl-long)
Copied to clipboard
| Challenge: | Graph Neural Networks (GNNs) with CLIP pipeline are difficult because of the scarcity of labeled data and text supervision, different levels of downstream tasks, and conceptual gaps between domains. |
| Approach: | They propose a multi-modal prompt learning paradigm to adapt pre-trained GNNs to downstream tasks with weak text supervision. |
| Outcome: | The proposed model can generalize graphs to unseen classes with weak text supervision. |
Towards Enhanced Immersion and Agency for LLM-based Interactive Drama (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have focused on the role of immersion and agency in interactive drama. |
| Approach: | They propose a playwriting-guided generation method that helps LLMs craft dramatic stories with substantially improved structures and narrative quality. |
| Outcome: | The proposed method improves storytelling quality and immersion and agency, while allowing agents to refine their reactions to align with the player’s intentions. |
Disambiguating Reference in Visually Grounded Dialogues through Joint Modeling of Textual and Multimodal Semantic Structures (2025.acl-long)
Copied to clipboard
| Challenge: | incorporating textual and multimodal reference resolution improves performance in visual-based reference resolution . Phrase grounding is a well-established task for understanding semantic relations between mentions and objects . ambiguities caused by pronouns and ellipses can arise in visually grounded dialogues . |
| Approach: | They propose a framework that unifies textual and multimodal reference resolution by mapping mention embeddings to object embeddements and selecting mentions or objects based on their similarity. |
| Outcome: | The proposed framework performs better in phrase grounding than other models for this task. |
Improving Factuality with Explicit Working Memory (2025.acl-long)
Copied to clipboard
Mingda Chen, Yang Li, Karthik Padthe, Rulin Shao, Alicia Yi Sun, Luke Zettlemoyer, Gargi Ghosh, Wen-tau Yih
| Challenge: | Large language models can generate factually inaccurate content, a problem known as hallucination. |
| Approach: | They propose an approach that integrates a working memory that receives feedback from external resources. |
| Outcome: | The proposed method outperforms baselines on four fact-seeking datasets and increases the factuality metric by 2 to 6 points absolute. |
Gradient-Adaptive Policy Optimization: Towards Multi-Objective Alignment of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Reinforcement learning from human feedback (RLHF) has emerged as a powerful technique for aligning large language models (LLMs) with human preferences. |
| Approach: | They propose a novel algorithm that uses multiple-gradient descent to optimize LLMs with diverse preferences to maximize trade-offs between objectives. |
| Outcome: | The proposed approach incorporates user preferences across different objectives and achieves Pareto solutions that better align with the user’s specific needs. |
Dynamic Parallel Tree Search for Efficient LLM Reasoning (2025.acl-long)
Copied to clipboard
Yifu Ding, Wentao Jiang, Shunyu Liu, Yongcheng Jing, Jinyang Guo, Yingjie Wang, Jing Zhang, Zengmao Wang, Ziwei Liu, Bo Du, Xianglong Liu, Dacheng Tao
| Challenge: | Recent methods focus on search accuracy while overlooking computational efficiency. |
| Approach: | They propose a parallelism framework that dynamically optimizes reasoning path in inference. |
| Outcome: | The proposed framework improves efficiency by 2-4 on average while maintaining or even surpassing existing reasoning algorithms in accuracy. |
Pre3: Enabling Deterministic Pushdown Automata for Faster Structured LLM Generation (2025.acl-long)
Copied to clipboard
Junyi Chen, Shihao Bai, Zaijun Wang, Siyu Wu, Chuheng Du, Hailong Yang, Ruihao Gong, Shengzhong Liu, Fan Wu, Guihai Chen
| Challenge: | Existing methods for structured generation of outputs are inefficient under large inference batches. |
| Approach: | They propose a new LLM-based method that parses LR(1) grammars into a pushdown automaton and exploits deterministic pushdown automation to optimize the constrained LLM decoding efficiency. |
| Outcome: | The proposed method improves time per output token (TPOT) by 40% and throughput by 36% . |
SHARE: An SLM-based Hierarchical Action CorREction Assistant for Text-to-SQL (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to self-correct text-to-SQL fail to demonstrate underlying reasoning path . authors propose **SHARE**, a self-revolution assistant for text-based error correction . |
| Approach: | They propose a "SHARE" assistant that enables LLMs to perform more precise error localization and efficient correction. |
| Outcome: | The proposed assistant performs more precise error localization and efficient correction for monolithic SQL queries. |
GenderAlign: An Alignment Dataset for Mitigating Gender Bias in Large Language Models (2025.acl-long)
Copied to clipboard
Tao Zhang, Ziqian Zeng, YuxiangXiao YuxiangXiao, Huiping Zhuang, Cen Chen, James R. Foulds, Shimei Pan
| Challenge: | Large Language Models (LLMs) generate content that exhibits gender biases, raising ethical concerns. |
| Approach: | They propose to use a dataset to identify gender biases in Large Language Models (LLMs) this dataset is a "chosen" and "rejected" LLM alignment is an effective approach to mitigate gender bias. |
| Outcome: | The proposed dataset shows that it reduces gender bias and improves quality. |
Large Language and Protein Assistant for Protein-Protein Interactions Prediction (2025.acl-long)
Copied to clipboard
Peng Zhou, Pengsen Ma, Jianmin Wang, Xibao Cai, Haitao Huang, Wei Liu, Longyue Wang, Lai Hou Tim, Xiangxiang Zeng
| Challenge: | Existing methods for predicting protein-protein interactions oversimplify the problem of PPI prediction in a semi-supervised manner. |
| Approach: | They propose a multimodal large language model that integrates proteins and PPI networks. |
| Outcome: | Experiments show that LLaPA can predict protein-protein interactions (mPPI) types and affinities based on sequence data. |
An Empirical Study of Many-to-Many Summarization with Large Language Models (2025.acl-long)
Copied to clipboard
Jiaan Wang, Fandong Meng, Zengkui Sun, Yunlong Liang, Yuxuan Cao, Jiarong Xu, Haoxiang Shi, Jie Zhou
| Challenge: | Recent studies have shown that large language models (LLMs) have strong multilingual abilities, giving them the potential to perform M2MS in real applications. |
| Approach: | They propose to use many-to-many summarization (M2MS) to generate a brief summary in any language given a document also in any other language. |
| Outcome: | The proposed model outperforms zero-shot LLMs in terms of automatic evaluations. |
Locate-and-Focus: Enhancing Terminology Translation in Speech Language Models (2025.acl-long)
Copied to clipboard
Suhang Wu, Jialong Tang, Chengyi Yang, Pei Zhang, Baosong Yang, Junhui Li, Junfeng Yao, Min Zhang, Jinsong Su
| Challenge: | Existing methods for terminology translation struggle with interference from irrelevant noise. |
| Approach: | They propose a Locate-and-Focus method that locates terminologies within utterances to construct translation knowledge by minimizing irrelevant information for ST models. |
| Outcome: | The proposed method locates terminologies within utterances and enhances the success rate of terminology translation while maintaining robust general translation performance. |
GuideBench: Benchmarking Domain-Oriented Guideline Following for LLM Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have been widely deployed as autonomous agents capable of following user instructions and making decisions in real-world applications. |
| Approach: | They propose a benchmark to evaluate LLMs' ability to follow domain-oriented guidelines . they evaluate Lms on three critical aspects: adherence to diverse rules, robustness to rule updates . |
| Outcome: | The proposed benchmark evaluates LLMs on three critical aspects: adherence to diverse rules, robustness to rule updates, and alignment with human preferences. |
TC–RAG: Turing–Complete RAG’s Case study on Medical LLM Systems (2025.acl-long)
Copied to clipboard
Xinke Jiang, Yue Fang, Rihong Qiu, Haoyu Zhang, Yongxin Xu, Hao Chen, Wentao Zhang, Ruizhe Zhang, Yuchen Fang, Xinyu Ma, Xu Chu, Junfeng Zhao, Yasha Wang
| Challenge: | Existing approaches to RAG neglect system state variables, resulting in poor performance and erroneous knowledge accumulation. |
| Approach: | They propose a framework that incorporates a Turing Complete System to manage state variables and manage retrieval halting. |
| Outcome: | The proposed framework improves on seven real-world healthcare datasets and shows that it is more accurate than existing methods. |
SoRFT: Issue Resolving with Subtask-oriented Reinforced Fine-Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing issue-resolving frameworks rely on commercial models, leading to high costs and privacy concerns. |
| Approach: | They propose a training approach to enhance issue resolving capability of LLMs by decomposing issue reasolving into subtasks. |
| Outcome: | The proposed approach improves issue-resolving performance and generalizes model . it is cost-effective and provides a cost-efficient alternative to commercial models . |
MiniLongBench: The Low-cost Long Context Understanding Benchmark for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing LCU benchmarks for large language models often result in prohibitively high evaluation costs . existing benchmarks exhibit significant redundancy, which means inefficiency in evaluation . |
| Approach: | They propose a data compression method tailored for long-text data with sparse information characteristics. |
| Outcome: | The proposed method reduces evaluation costs to 4.5% of the long-text benchmark LongBench . the proposed method is based on a long-term LCU benchmark with sparse information characteristics . |
Divide-Then-Align: Honest Alignment based on the Knowledge Boundary of RAG (2025.acl-long)
Copied to clipboard
Xin Sun, Jianan Xie, Zhongqi Chen, Qiang Liu, Shu Wu, Yuehe Chen, Bowen Song, Zilei Wang, Weiqiang Wang, Liang Wang
| Challenge: | Large language models (LLMs) augmented with retrieval systems have significantly advanced natural language processing tasks by integrating external knowledge sources. |
| Approach: | They propose a method that conditions large language models to generate answers even in the absence of reliable knowledge. |
| Outcome: | The proposed approach balances accuracy with appropriate abstention, enhancing the reliability and trustworthiness of retrieval-augmented systems. |
PwnGPT: Automatic Exploit Generation Based on Large Language Models (2025.acl-long)
Copied to clipboard
Wanzong Peng, Lin Ye, Xuetao Du, Hongli Zhang, Dongyang Zhan, Yunting Zhang, Yicheng Guo, Chen Zhang
| Challenge: | Automated exploit generation (AEG) is the automatic discovery and exploitation of vulnerabilities against unknown targets. |
| Approach: | They propose an automatic exploit generation framework that automatically solves pwn challenges by using large language models. |
| Outcome: | The proposed framework improves the completion rate of exploits on the openAI o1-preview model and the GPT-4o model. |
VMLU Benchmarks: A comprehensive benchmark toolkit for Vietnamese LLMs (2025.acl-long)
Copied to clipboard
Cuc Thi Bui, Nguyen Truong Son, Truong Van Trang, Lam Viet Phung, Pham Nhut Huy, Hoang Anh Le, Quoc Huu Van, Phong Nguyen-Thuan Do, Van Le Tran Truc, Duc Thanh Chau, Le-Minh Nguyen
| Challenge: | The evolution of Large Language Models (LLMs) has underscored the need for benchmarks designed for various languages and cultural contexts. |
| Approach: | They propose to use Vietnamese multitask language understanding (VMLU) benchmarks to assess different capabilities of LLMs, including general knowledge, reading comprehension, reasoning, and conversational skills. |
| Outcome: | The VMLU Benchmarks assess LLMs' general knowledge, reading comprehension, reasoning, and conversational skills. |
Scaling up the State Size of RNN LLMs for Long-Context Scenarios (2025.acl-long)
Copied to clipboard
| Challenge: | Existing RNN-based LLMs struggle with long-context scenarios due to their quadratic computational complexity and linear memory requirements. |
| Approach: | They propose an efficient scaling method to scale RNN models to match the 2k context length of Transformers with small parameters overhead. |
| Outcome: | The proposed method improves long-context understanding and improves performance on FDA recall-intensive tasks. |
Unifying Continuous and Discrete Text Diffusion with Non-simultaneous Diffusion Processes (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results demonstrate NeoDiff’s superior performance compared to baselines of non-autoregressive continuous and discrete diffusion models, iterative-based methods and autoregressive diffusion-based approaches. |
| Approach: | They propose a discrete and continuous diffusion model that integrates the strengths of discrete, continuous and continuous approaches. |
| Outcome: | The proposed model unifies the theories of discrete and continuous diffusion models, offering a more principled and effective framework for text generation. |
A Strategic Coordination Framework of Small LMs Matches Large LMs in Data Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models suffer from high computational costs and environmental inefficiency . smaller LMs are more accessible and sustainable, but their individual capabilities often fall short . a collaborative framework for small LM combines specialized roles to iterative refinement and quality control . |
| Approach: | They propose a framework that aggregates specialized roles across small LMs to iterative refinement and quality control typically achieved by a single large LM. |
| Outcome: | The proposed framework aggregates specialized roles across small LMs to iterative refinement and quality control typically achieved by large LM. |
Defining and Evaluating Visual Language Models’ Basic Spatial Abilities: A Perspective from Psychometrics (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies assessing the spatial abilities of VLMs lack a solid theoretical foundation and lack measurable data. |
| Approach: | They propose a psychometric framework defining five basic spatial abilities in Visual Language Models. |
| Outcome: | The proposed framework defines five basic spatial abilities in Visual Language Models (VLMs) it provides a comprehensive evaluation benchmark and methodological perspective for embodied AI development . |
SPHERE: Unveiling Spatial Blind Spots in Vision-Language Models Through Hierarchical Evaluation (2025.acl-long)
Copied to clipboard
Wenyu Zhang, Wei En Ng, Lixin Ma, Yuwen Wang, Junqi Zhao, Allison Koenecke, Boyang Li, Wanglu Wanglu
| Challenge: | Current vision-language models lack multi-dimensional spatial reasoning capabilities for human-like understanding and applications. |
| Approach: | They propose a hierarchical evaluation framework that probes models across increasing levels of complexity and integrates spatial, visual, and logical understanding. |
| Outcome: | The proposed framework probes models across increasing levels of complexity, from basic skills to multi-skill integration and high-level reasoning that combines spatial, visual, and logical understanding. |
User-side Model Consistency Monitoring for Open Source Large Language Models Inference Services (2025.acl-long)
Copied to clipboard
| Challenge: | Open-source large language models have attracted substantial user base by offering quality comparable to closed-source models at a significantly lower cost. |
| Approach: | They propose a paradigm for model consistency monitoring on the user side that constructs metrics based on logits produced by LLMs to differentiate sequences generated by degraded models. |
| Outcome: | The proposed method is implementable on consumer-grade devices and demonstrates that it is cost-effective and reliable. |
Jailbreaking? One Step Is Enough! (2025.acl-long)
Copied to clipboard
Weixiong Zheng, Peijian Zeng, YiWei Li, Hongyan Wu, Nankai Lin, Junhao Chen, Aimin Yang, Yongmei Zhou
| Challenge: | Large language models (LLMs) excel in various tasks but remain vulnerable to jailbreak attacks, where adversaries manipulate prompts to generate harmful outputs. |
| Approach: | They propose a Reverse Embedded Defense Attack mechanism that disguises the attack intention as the "defense" intention against harmful content. |
| Outcome: | The proposed method outperforms existing methods on open-source and closed-source models and enables successful jailbreak in one iteration. |
Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning (2025.acl-long)
Copied to clipboard
Yongxin Xu, Ruizhe Zhang, Xinke Jiang, Yujie Feng, Yuzhen Xiao, Xinyu Ma, Runchuan Zhu, Xu Chu, Junfeng Zhao, Yasha Wang
| Challenge: | Existing methods for integrating internal and external knowledge lack effective control mechanisms for generating hallucinations and dealing with outdated knowledge. |
| Approach: | They propose a framework that decouples, identifies, and purposefully optimizes parameter subspaces related to adherence and robustness. |
| Outcome: | The proposed framework decouples, identifies, and purposefully optimizes parameter subspaces related to adherence and robustness. |
PaSa: An LLM Agent for Comprehensive Academic Paper Search (2025.acl-long)
Copied to clipboard
| Challenge: | We introduce PaSa, an advanced Paper Search agent powered by large language models . despite being trained on synthetic data, PaSA outperforms existing baselines on RealScholarQuery . |
| Approach: | They introduce PaSa, an advanced Paper Search agent powered by large language models . they optimize PaSA using a synthetic dataset, AutoScholarQuery, which includes 35k fine-grained queries . |
| Outcome: | The paper analyzes the performance of a paper search agent using a synthetic dataset . it significantly outperforms existing benchmarks on RealScholarQuery . |
Less Mature is More Adaptable for Sentence-level Language Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies fine-tune encoders or contrastive learning approaches to learn sentence representations. |
| Approach: | They propose to use sentence-level models to study how sentence representations influence downstream task performance. |
| Outcome: | The proposed models outperform token-level models in terms of time and data efficiency. |
EpMAN: Episodic Memory AttentioN for Generalizing to Longer Contexts (2025.acl-long)
Copied to clipboard
Subhajit Chaudhury, Payel Das, Sarathkrishna Swaminathan, Georgios Kollias, Elliot Nelson, Khushbu Pahwa, Tejaswini Pedapati, Igor Melnyk, Matthew Riemer
| Challenge: | Recent advances in Large Language Models (LLMs) have yielded impressive successes on many language tasks, but efficient processing of long contexts remains a significant challenge. |
| Approach: | They propose a method for processing long contexts in an episodic memory module while holistically attending to semantically-relevant context chunks. |
| Outcome: | The proposed method outperforms baseline decoders on multiple long-context recall and question-answering benchmarks on 16k to 256k tokens. |
UORA: Uniform Orthogonal Reinitialization Adaptation in Parameter Efficient Fine-Tuning of Large Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods such as LoRA and VeRA use a low-rank approximation method that reduces the number of trainable parameters without compromising performance. |
| Approach: | They propose a parameter-efficient fine-tuning approach that leverages a low-rank approximation method that reduces the number of trainable parameters without compromising performance. |
| Outcome: | The proposed approach outperforms existing methods on GLUE and E2E benchmarks and is effective in instruction-tuning large language models and image classification models. |
Agri-CM3: A Chinese Massive Multi-modal, Multi-level Benchmark for Agricultural Understanding and Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks lack comprehensive evaluations, particularly in multi-level reasoning, making it difficult to identify model limitations. |
| Approach: | They propose to use Agri-CM3 to assess multi-level reasoning in agricultural management by integrating multiple data modalities. |
| Outcome: | The Agri-CM3 benchmark includes 3,939 images and 15,901 multi-level multiple-choice questions with detailed explanations. |
TROVE: A Challenge for Fine-Grained Text Provenance via Source Sentence Tracing and Relationship Classification (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models have demonstrated great potential in natural language generation, but their widespread adoption has raised concerns regarding content reliability and accountability. |
| Approach: | They propose a challenge to trace each sentence of a target text back to specific source sentences within potentially lengthy or multi-document inputs. |
| Outcome: | The proposed challenge traces each sentence of a target text back to specific source sentences . the dataset includes 11 scenarios covering QA and summarization in english and Chinese . |
CaLMQA: Exploring culturally specific long-form question answering across 23 languages (2025.acl-long)
Copied to clipboard
| Challenge: | Despite rising global usage of large language models, their ability to generate *long-form* answers to *culturally specific* questions remains unexplored in many languages. |
| Approach: | They perform the first study of textual multilingual long-form QA by creating a dataset of culturally specific questions across 23 different languages. |
| Outcome: | The results show that the best models make critical surface-level errors for many languages and their understanding of diverse cultures. |
Croppable Knowledge Graph Embedding (2025.acl-long)
Copied to clipboard
| Challenge: | Knowledge Graph Embedding (KGE) is a common approach for Knowledge Grasse (KGs) in AI tasks. |
| Approach: | They propose a new KGE training framework MED that allows one training to obtain a croppable KGE model for multiple scenarios with different dimensional needs. |
| Outcome: | The proposed framework improves low-dimensional sub-models and makes high-dimensional models retain the low-dimension sub-modells’ capacity. |
HyKGE: A Hypothesis Knowledge Graph Enhanced RAG Framework for Accurate and Reliable Medical LLMs Responses (2025.acl-long)
Copied to clipboard
Xinke Jiang, Ruizhe Zhang, Yongxin Xu, Rihong Qiu, Yue Fang, Zhiyuan Wang, Jinyi Tang, Hongxin Ding, Xu Chu, Junfeng Zhao, Yasha Wang
| Challenge: | Recent approaches suffer from insufficient and repetitive knowledge retrieval, tedious and time-consuming query parsing, and monotonous knowledge utilization. |
| Approach: | They propose a retrieval-augmented generation framework which leverages LLMs’ powerful reasoning capacity to compensate for the incompleteness of user queries. |
| Outcome: | The proposed framework improves the accuracy and reliability of Large Language Models (LLMs) by combining the rich knowledge of LLMs with Hypothesis Outputs. |
LongRecipe: Recipe for Efficient Long Context Generalization in Large Language Models (2025.acl-long)
Copied to clipboard
Zhiyuan Hu, Yuliang Liu, Jinman Zhao, Suyuchen Wang, WangYan WangYan, Wei Shen, Qing Gu, Anh Tuan Luu, See-Kiong Ng, Zhiwei Jiang, Bryan Hooi
| Challenge: | Large language models face significant challenges in handling long-context tasks because of their limited effective context window size during pretraining, which restricts their ability to generalize over extended sequences. |
| Approach: | They propose a training strategy for extending the context window of LLMs including impactful token analysis, position index transformation, and training optimization strategies. |
| Outcome: | Experiments on three types of LLMs show that LongRecipe can utilize long sequences while requiring only 30% of the target context window size. |
BeamLoRA: Beam-Constraint Low-Rank Adaptation (2025.acl-long)
Copied to clipboard
Naibin Gu, Zhenyu Zhang, Xiyu Liu, Peng Fu, Zheng Lin, Shuohuan Wang, Yu Sun, Hua Wu, Weiping Wang, Haifeng Wang
| Challenge: | Low-Rank Adaptation (LoRA) is one of the most efficient parameter-efficient fine-tuning methods. |
| Approach: | They propose to conceptualize each LoRA module as a beam where each rank corresponds to a potential sub-solution. |
| Outcome: | The proposed method improves performance on three base models and 12 datasets. |
GODBench: A Benchmark for Multimodal Large Language Models in Video Comment Art (2025.acl-long)
Copied to clipboard
Yiming Lei, Chenkai Zhang, Zeming Liu, Haitao Leng, ShaoGuo Liu, Tingting Gao, Qingjie Liu, Yunhong Wang
| Challenge: | Existing benchmarks for video comment art are constrained by their limited modalities and insufficient categories, hindering creativity in video-based comment art creation. |
| Approach: | They propose a benchmark that integrates video and text modalities to evaluate MLLMs’ abilities to compose video Comment art. |
| Outcome: | The proposed framework integrates video and text modalities to evaluate MLLMs’ abilities to compose video comment art. |
UniLR: Unleashing the Power of LLMs on Multiple Legal Tasks with a Unified Legal Retriever (2025.acl-long)
Copied to clipboard
Ang Li, Yiquan Wu, Yifei Liu, Ming Cai, Lizhi Qing, Shihang Wang, Yangyang Kang, Chengyuan Liu, Fei Wu, Kun Kuang
| Challenge: | Existing retrieval methods are designed for general domains, struggling with legal knowledge, or tailored for specific legal tasks, unable to handle diverse legal knowledge types. |
| Approach: | They propose a novel retrieval method that integrates specialized knowledge into LLMs. |
| Outcome: | The proposed method can perform multiple legal retrieval tasks for LLMs. |
Generative Psycho-Lexical Approach for Constructing Value Systems in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have raised concerns regarding their intrinsic values. |
| Approach: | They propose a psychologically grounded five-factor value system for Large Language Models that integrates psychological principles with cutting-edge AI priorities. |
| Outcome: | The proposed value system meets standard psychological criteria, improves LLM safety prediction, and enhances Llm alignment, when compared to the canonical Schwartz’s values. |
Beyond Dialogue: A Profile-Dialogue Alignment Framework Towards General Role-Playing Language Model (2025.acl-long)
Copied to clipboard
| Challenge: | Existing role-playing training methods often lack profile-dialogue alignment at the sentence level. |
| Approach: | They propose a framework that aligns dialogue with profile traits for each scenario, eliminating biases during training. |
| Outcome: | The proposed model outperforms most proprietary role-playing models and is fully automated and low-cost. |
ACECODER: Acing Coder RL via Automated Test-Case Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Recent coder models have been driven by supervised fine-tuning, but the potential of reinforcement learning remains unexplored due to the lack of reliable reward data/model in the code domain. |
| Approach: | They propose a pipeline that generates extensive test-case pairs from existing code data and constructs preference pairs based on pass rates over sampled programs. |
| Outcome: | The proposed pipeline generates extensive (question, test-cases) pairs from existing code data and trains them with Bradley-Terry loss. |
Quantifying Semantic Emergence in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods for large language models (LLMs) focus on coarse-grained text, not providing interpretations for the behavior of finergrained tokens. |
| Approach: | They propose a quantitative metric to measure large language models’ ability to extract semantics from input tokens. |
| Outcome: | The proposed metric compares the entropy reduction observed for a sequence of tokens and individual tokens. |
DebateCoder: Towards Collective Intelligence of LLMs via Test Case Driven LLM Debate for Code Generation (2025.acl-long)
Copied to clipboard
Jizheng Chen, Kounianhua Du, Xinyi Dai, Weiming Zhang, Xihuai Wang, Yasheng Wang, Ruiming Tang, Weinan Zhang, Yong Yu
| Challenge: | Existing debate-based approaches to code generation are limited due to several reasons: 1) Reliance on different instances of the same LLM for debate, 2) under-utilization of test cases, and 3) reliance on third-party moderators for result consolidation and decision-making. |
| Approach: | They propose to use test cases to analyze code and identify bugs while opposing models generate test cases for each other to challenge each other's code during the debate process. |
| Outcome: | The proposed model collects intelligence of LLMs via test case-driven debate for code generation. |
The Tug of War Within: Mitigating the Fairness-Privacy Conflicts in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to enhance an LLM's privacy awareness with thousands of samples decrease its fairness awareness. |
| Approach: | They propose a training-free method to Suppress the Privacy and faIrness coupled Neurons (SPIN) which theoretically and empirically decreases the mutual information between fairness and privacy awareness. |
| Outcome: | The proposed method reduces the mutual information between fairness and privacy awareness without compromising general capabilities. |
GraphInsight: Unlocking Insights in Large Language Models for Graph Structure Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models struggle with comprehending graphical structure information through prompts of graph description sequences, especially as the graph size increases. |
| Approach: | They propose a framework to improve LLMs’ comprehension of both macro- and micro-level graphical information by placing critical graphical data in positions where LLM's exhibit stronger memory performance. |
| Outcome: | The proposed framework outperforms all other graph description methods in understanding graph structures of varying sizes. |
Phonotomizer: A Compact, Unsupervised, Online Training Approach to Real-Time, Multilingual Phonetic Segmentation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to phonetic segmentation are hierarchical and end-to-end . many mistakes in final output stem from subtle segmenter perturbations . |
| Approach: | They propose a phonetic segmentation system that trains on raw sound files alone . it can modulate computational exactness and reduce acoustic model size, they argue . |
| Outcome: | The proposed method reduces the size of the acoustic model and training epochs. |
A Multi-persona Framework for Argument Quality Assessment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for argument quality assessment do not consider multi-perspective evaluation due to subjective nature of arguments. |
| Approach: | They propose a multi-persona framework for argument quality assessment that simulates diverse evaluator perspectives through large language models. |
| Outcome: | The proposed framework outperforms baselines while providing comprehensive multi-perspective rationales on IBM-Rank-30k and IBM-ArgQ-5.3kArgs datasets. |
Safe: Enhancing Mathematical Reasoning in Large Language Models via Retrospective Step-aware Formal Verification (2025.acl-long)
Copied to clipboard
Chengwu Liu, Ye Yuan, Yichun Yin, Yan Xu, Xin Xu, Zaoyu Chen, Yasheng Wang, Lifeng Shang, Qun Liu, Ming Zhang
| Challenge: | Chain-of-Thought prompting is a de facto method to elicit reasoning capabilities from large language models (LLMs). |
| Approach: | They propose a step-aware formal verification framework Safe to address hallucinations in CoT prompting . they propose 'formal step' as a benchmark for step correctness theorem proving with 30,809 formal statements. |
| Outcome: | The proposed framework shows significant performance improvement while offering interpretable and verifiable evidence. |
SAM Decoding: Speculative Decoding via Suffix Automaton (2025.acl-long)
Copied to clipboard
| Challenge: | Speculative decoding (SD) methods are inefficient and rely on single retrieval resources. |
| Approach: | They propose a retrieval-based speculative decoding method that adapts the suffix automaton for efficient draft generation by utilizing the generating text sequence and static text corpus. |
| Outcome: | The proposed method can find the longest suffix match and can be integrated with existing methods to generalize to broader domains. |
PsyAdvisor: A Plug-and-Play Strategy Advice Planner with Proactive Questioning in Psychological Conversations (2025.acl-long)
Copied to clipboard
| Challenge: | Current psychological LLMs are constrained by passive response mechanisms, limiting their capacity to deploy proactive strategies for psychological counseling. |
| Approach: | They propose a dataset that provides a multi-turn conversation dataset with interpretive labels including strategy decision logic and reaction attribution. |
| Outcome: | The proposed model significantly improves proactive questioning capacity, conversation depth, and response quality. |
HomeBench: Evaluating LLMs in Smart Homes with Valid and Invalid Instructions Across Single and Multiple Devices (2025.acl-long)
Copied to clipboard
| Challenge: | Existing state-of-the-art LLMs cannot perform well in situations where instructions are invalid or multiple devices are involved. |
| Approach: | They propose to integrate large language models into smart home assistants by enhancing their ability to accurately understand user needs and respond appropriately. |
| Outcome: | The proposed dataset is the first with valid and invalid instructions across devices . it achieves only 0.0% success rate in the scenario of invalid multi-device instructions . |
Advancing Zero-shot Text-to-Speech Intelligibility across Diverse Domains via Preference Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing zero-shot text-to-speech systems struggle in challenging scenarios such as tongue twisters, repeated words, code-switching, and cross-lingual synthesis. |
| Approach: | They propose a dataset that leverages preference alignment techniques to improve performance . they also extend the Direct Preference Optimization framework to accommodate diverse TTS architectures . |
| Outcome: | The proposed dataset improves intelligibility, similarity, and audio quality for multiple models across domains. |
GiFT: Gibbs Fine-Tuning for Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Training Large Language Models (LLMs) with synthetic data is a prevalent practice in code generation. |
| Approach: | They propose a method to fine-tune large language models with code drawn from a conditional distribution, conditioned on a specific seed description. |
| Outcome: | The proposed method improves performance on four datasets and shows that it can be used to fine-tune LLMs with code derived from the marginal distribution. |
Enhancing Interpretable Image Classification Through LLM Agents and Conditional Concept Bottleneck Models (2025.acl-long)
Copied to clipboard
| Challenge: | Concept Bottleneck Models (CBMs) map visual representations to a set of humanunderstandable textual concepts, which are then interpreted by a linear combination of these concept scores. |
| Approach: | They propose a dynamic, agent-based approach that adjusts the concept bank in response to environmental feedback, optimizing the number of concepts for sufficiency yet concise coverage. |
| Outcome: | The proposed model improves classification accuracy by 6% and interpretability assessments by 30%. |
Reliably Bounding False Positives: A Zero-Shot Machine-Generated Text Detection Framework via Multiscaled Conformal Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods focus excessively on detection accuracy, neglecting the societal risks posed by high false positive rates (FPRs). |
| Approach: | They propose a Conformal Prediction framework that constrains the upper bound of false positive rates and introduces a real-time detection framework. |
| Outcome: | The proposed framework reduces false positive rates and improves detection performance. |
RSCF: Relation-Semantics Consistent Filter for Entity Embedding of Knowledge Graph (2025.acl-long)
Copied to clipboard
| Challenge: | Knowledge graph embeddings suffer from incompleteness, a problem that is often overlooked . a generalized plug-in approach to SFBR disrupts consistency by concentrating embeddables under entity-based regularization . |
| Approach: | They propose a plug-in KGE method that uses relation specific entity transformation to enhance semantic consistency. |
| Outcome: | The proposed method outperforms state-of-the-art methods in knowledge graph embedding tasks . the proposed method is based on a plug-in approach that disrupts consistency . |
RolePlot: A Systematic Framework for Evaluating and Enhancing the Plot-Progression Capabilities of Role-Playing Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research has focused on role-playing agents’ ability to portray specified characters, but their ability to advance the plot requires substantial improvements to deliver more engaging interaction. |
| Approach: | They propose a role-playing framework to evaluate and enhance the plot-progression capabilities of role-players. |
| Outcome: | The proposed framework improves RPAs’ ability to time plot developments and yields a significant increase in conversation turns and sustained higher arousal levels. |
TreeRL: LLM Reinforcement Learning with On-Policy Tree Search (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for On-Policy LLM RL typically train a separate process reward model, which suffers from distribution mismatch and reward hacking. |
| Approach: | They propose a reinforcement learning framework that directly incorporates on-policy tree search for RL training. |
| Outcome: | Experiments on math and code reasoning benchmarks show that tree search achieves superior performance compared to traditional ChainRL. |
Can a Single Model Master Both Multi-turn Conversations and Tool Use? CoALM: A Unified Conversational Agentic Language Model (2025.acl-long)
Copied to clipboard
Emre Can Acikgoz, Jeremiah Greer, Akul Datta, Ze Yang, William Zeng, Oussama Elachqar, Emmanouil Koukoumidis, Dilek Hakkani-Tür, Gokhan Tur
| Challenge: | Large Language Models (LLMs) with API-calling capabilities enabled building effective Language Agents (LA) current approaches excel in one domain but underperform in the other. |
| Approach: | They propose a unified approach that integrates both conversational and agentic capabilities. |
| Outcome: | The proposed model outperforms top domain-specific models across three benchmarks. |
Single-to-mix Modality Alignment with Multimodal Large Language Model for Document Image Machine Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Document Image Machine Translation (DIMT) faces generalization challenges due to limited training data and the complex interplay between visual and textual information. |
| Approach: | They propose a single-to-mix Modality alignment framework leveraging Multimodal Large Language Models (MLLMs) this framework aligns an imageonly encoder with multimodal representations of an MLLM pre-trained on large-scale document image datasets. |
| Outcome: | The proposed framework improves translation quality in cross-domain generalization and challenging document image scenarios. |
SDPO: Segment-Level Direct Preference Optimization for Social Agents (2025.acl-long)
Copied to clipboard
Aobo Kong, Wentao Ma, Shiwan Zhao, Yongbin Li, Yuchuan Wu, Ke Wang, Xiaoqian Liu, Qicheng Li, Yong Qin, Fei Huang
| Challenge: | Direct Preference Optimization (DPO) has proven effective in aligning LLM behavior with human preferences across various tasks, but is limited in multi-turn social interactions. |
| Approach: | They propose a method which dynamically selects key segments within interactions to optimize multi-turn agent behavior. |
| Outcome: | The proposed methods outperform existing methods and proprietary LLMs on the SOTOPIA benchmark and show that they can improve social intelligence. |
KokoroChat: A Japanese Psychological Counseling Dialogue Dataset Collected via Role-Playing by Trained Counselors (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have explored using large language models to augment counseling dialogue datasets, but data from real-world counseling environments may suffer from limited diversity and authenticity. |
| Approach: | They propose to use a Japanese psychological counseling dialogue dataset to simulate counselor-client interactions by using open-source LLMs. |
| Outcome: | The proposed model improves the quality of generated counseling responses and the automatic evaluation of counseling dialogues. |
SURVEYFORGE : On the Outline Heuristics, Memory-Driven Generation, and Multi-dimensional Evaluation for Automated Survey Writing (2025.acl-long)
Copied to clipboard
| Challenge: | SURVEYFORGE automates survey paper writing, but quality gap between LLM-generated and human-written surveys remains significant. |
| Approach: | They propose a survey tool that automatically generates and refines human-written surveys. |
| Outcome: | Experiments show that SURVEYFORGE outperforms previous work such as AutoSurvey in outline quality and content quality. |
Making LLMs Better Many-to-Many Speech-to-Text Translators with Curriculum Learning (2025.acl-long)
Copied to clipboard
Yexing Du, Youcheng Pan, Ziyang Ma, Bo Yang, Yifan Yang, Keqi Deng, Xie Chen, Yang Xiang, Ming Liu, Bing Qin
| Challenge: | Existing studies on English-centric translation tasks have focused on multimodal large language models, but the exploration of many-to-many translation is limited by the scarcity of parallel data. |
| Approach: | They propose a three-stage curriculum learning strategy that leverages the machine translation capabilities of large language models and adapts them to S2TT tasks. |
| Outcome: | The proposed strategy achieves state-of-the-art average performance in 1514 language pairs, requiring fewer than 10 hours of speech data per language to achieve competitive results. |
AbGen: Evaluating Large Language Models in Ablation Study Design and Evaluation for Scientific Research (2025.acl-long)
Copied to clipboard
Yilun Zhao, Weiyuan Chen, Zhijian Xu, Manasi Patwardhan, Chengye Wang, Yixin Liu, Lovekesh Vig, Arman Cohan
| Challenge: | a benchmark designed to evaluate the capabilities of LLMs in designing ablation studies for scientific research is available online. |
| Approach: | They propose to use a benchmark to evaluate LLMs' ability to design ablation studies . they investigate whether current automated evaluation methods are not reliable . |
| Outcome: | The benchmark compared leading LLMs with human experts on generating detailed ablation study designs . the results show that current evaluation methods are not reliable for the task . |
Redundancy Principles for MLLMs Benchmarks (2025.acl-long)
Copied to clipboard
Zicheng Zhang, Xiangyu Zhao, Xinyu Fang, Chunyi Li, Xiaohong Liu, Xiongkuo Min, Haodong Duan, Kai Chen, Guangtao Zhai
| Challenge: | Rapid growth of Multi-modality Large Language Models has led to significant redundancy among benchmarks. |
| Approach: | They propose a framework to improve MLLM benchmark design by identifying redundancy at three levels: dimension, instance, and cross-benchmark redundancies. |
| Outcome: | The proposed framework streamlines evaluations and enhances reliability. |
WavRAG: Audio-Integrated Retrieval Augmented Generation for Spoken Dialogue Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing RAG frameworks rely on Automatic Speech Recognition to process speech input, which discards crucial audio information and increases computational overhead. |
| Approach: | They propose a retrieval augmented generation framework with native, end-to-end audio support that integrates audio and text into a unified knowledge representation. |
| Outcome: | The proposed framework can perform 10x faster than current pipelines while delivering 10x acceleration. |
ChildMandarin: A Comprehensive Mandarin Speech Dataset for Young Children Aged 3-5 (2025.acl-long)
Copied to clipboard
Jiaming Zhou, Shiyao Wang, Shiwan Zhao, Jiabei He, Haoqin Sun, Hui Wang, Cheng Liu, Aobo Kong, Yujie Guo, Xi Yang, Yequan Wang, Yonghua Lin, Yong Qin
| Challenge: | Automatic speech recognition systems have advanced significantly with models like Whisper, Conformer, and self-supervised frameworks such as Wav2vec 2.0. |
| Approach: | They propose to use Mandarin speech datasets to analyze pronunciation and tone of children aged 3 to 5 and evaluate their models on speaker verification (SV) They find that the datasets are more robust than those used by adult speech recognition systems and are open-source and available for all academic purposes. |
| Outcome: | The proposed dataset includes 41.25 hours of speech with carefully crafted manual transcriptions, collected from 397 speakers across various provinces in China, with balanced gender representation. |
Finding the Sweet Spot: Preference Data Construction for Scaling Preference Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models generate unintended outputs due to their unsupervised nature. |
| Approach: | They propose a method to construct preference pairs of selected and rejected LLMs by repeated random sampling to improve alignment performance. |
| Outcome: | The proposed method improves performance as the sample size increases. |
Enhancing Safe and Controllable Protein Generation via Knowledge Preference Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Protein language models pose significant risks of generating harmful sequences, e.g., viral transmissibility, drug resistance, environmental imbalances, public health crises, etc. |
| Approach: | They propose a protein-based model that integrates prior knowledge via a Protein Safety Knowledge Graph to minimize the risk of generating harmful sequences. |
| Outcome: | The proposed framework reduces the likelihood of producing hazardous sequences while maintaining high functionality. |
SINCon: Mitigate LLM-Generated Malicious Message Injection Attack for Rumor Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods define important nodes as important and target them for attacks if the model treats nodes’ predictive influence more uniformly . Existing approaches target high predictive influence nodes but are vulnerable to malicious message injection attacks. |
| Approach: | They propose a defense mechanism that encourages the model to learn graph representations where nodes with varying importance have a more uniform influence on predictions. |
| Outcome: | Extensive experiments on the Twitter and Weibo datasets show that similarizing the predictive Influence of nodes with Contrastive Learning significantly enhances resistance against LLM-driven message injection attacks. |
Outlier-Safe Pre-Training for Robust 4-Bit Quantization of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Quantization is a practical solution for deploying Large Language Models in resource-constrained environments. |
| Approach: | They propose an outlier-safe pre-training approach that prevents outlier formation . they validate a 1.4B-parameter model on 1 trillion tokens with no outliers . |
| Outcome: | The proposed model achieves a 35.7 average score on 1 trillion tokens with 2% training overhead. |
Agentic Knowledgeable Self-awareness (2025.acl-long)
Copied to clipboard
Shuofei Qiao, Zhisong Qiu, Baochang Ren, Xiaobin Wang, Xiangyuan Ru, Ningyu Zhang, Xiang Chen, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen
| Challenge: | Large Language Models (LLMs) have achieved considerable performance across various agentic planning tasks. |
| Approach: | They propose a data-centric approach that applies agents with knowledgeable self-awareness like humans to a heuristic situation judgement criterion to mark special tokens on their self-explored trajectories for collecting training data. |
| Outcome: | The proposed paradigm outperforms baseline models on various tasks with minimal external knowledge. |
A Unified Agentic Framework for Evaluating Conditional Image Generation (2025.acl-long)
Copied to clipboard
Jifang Wang, Yangxue Yangxue, Longyue Wang, Zhenran Xu, Yiyu Wang, Yaowei Wang, Weihua Luo, Kaifu Zhang, Baotian Hu, Min Zhang
| Challenge: | Conditional image generation is a popular and personalization-oriented task, but there are challenges in developing task-agnostic, reliable, and explainable evaluation metrics. |
| Approach: | They propose a unified agentic framework for comprehensive evaluation of conditional image generation tasks. |
| Outcome: | The proposed framework achieves a high correlation with human assessments on seven prominent image generation tasks. |
Planning-Driven Programming: A Large Language Model Programming Workflow (2025.acl-long)
Copied to clipboard
| Challenge: | Recent research suggests continuous program refinements through visible tests to improve code generation accuracy in large language models (LLMs). |
| Approach: | They propose an LLM programming workflow to improve both initial code generation and subsequent refinements within a structured two-phase workflow. |
| Outcome: | The proposed workflow improves both initial code generation and subsequent refinements within a structured two-phase workflow. |
Can Knowledge Graphs Make Large Language Models More Trustworthy? An Empirical Study Over Open-ended Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks for integrating Knowledge Graphs with Large Language Models focus on closed-ended tasks, leaving a gap in evaluating performance on more complex, real-world scenarios. |
| Approach: | They propose a benchmark to evaluate LLMs augmented with KGs in open-ended, real-world question answering settings. |
| Outcome: | The proposed benchmark reflects practical complexities through diverse question types and incorporates metrics to quantify both hallucination rates and reasoning improvements in LLM+KG models. |
Nudging: Inference-time Alignment of LLMs via Guided Decoding (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) require alignment to effectively and safely follow user instructions. |
| Approach: | They propose a simple, training-free algorithm that aligns any base model at inference time using a small aligned model. |
| Outcome: | The proposed algorithm outperforms large aligned models on open-instruction tasks without training. |
Unveiling Attractor Cycles in Large Language Models: A Dynamical Systems View of Successive Paraphrasing (2025.acl-long)
Copied to clipboard
| Challenge: | Dynamical systems theory provides a framework for understanding iterative processes and evolution over time. |
| Approach: | They propose to apply this perspective to large language models which iteratively map input text to output text and re-express meaning with linguistic variation. |
| Outcome: | The proposed model reveals that paraphrases re-express meaning with linguistic variation limiting linguistic diversity . |
SCAR: Data Selection via Style Consistency-Aware Response Ranking for Efficient Instruction-Tuning of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies show that manually ensuring a consistent response style and maintaining high data quality can significantly improve the performance of fine-tuned Large Language Models (LLMs). |
| Approach: | They introduce a style-aware response ranking system that prioritizes instruction-response pairs based on their stylistic consistency. |
| Outcome: | The proposed model matches or surpasses models trained on the entire dataset in coding and open-ended question-answering benchmarks. |
HFT: Half Fine-Tuning for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) with one or more fine-tuning phases can unlock various capabilities, but can be catastrophic forgetting during sequential training. |
| Approach: | They propose a method to regularly reset partial parameters to mitigate forgetting issues by using half fine-tuning instead of full fine-uning. |
| Outcome: | The proposed approach reduces the risk of catastrophic forgetting during training and the parametric knowledge lost during training may be overwhelmed by incoming training data. |
Beyond Surface Simplicity: Revealing Hidden Reasoning Attributes for Precise Commonsense Diagnosis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing commonsense question answering benchmarks often treat these aspects in isolation, resulting in evaluation accuracy differences of up to 24.8% across different difficulty levels. |
| Approach: | They propose a framework that reveals hidden reasoning attributes behind commonsense questions by leveraging the knowledge generated during the reasoning process. |
| Outcome: | The proposed framework reveals hidden reasoning attributes behind commonsense questions by leveraging the knowledge generated during the reasoning process. |
From Objectives to Questions: A Planning-based Framework for Educational Mathematical Question Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Traditional generation methods focus primarily on textual quality, but they fail to meet complex, multifaceted educational requirements. |
| Approach: | They propose a method for automatic generating high-quality mathematical problems that align with educational objectives using a dataset of 16k mathematical questions with multi-dimensional educational objectives. |
| Outcome: | The proposed method improves generating high-quality mathematical questions that meet multi-dimensional educational objectives. |
RankCoT: Refining Knowledge for Retrieval-Augmented Generation through Ranking Chain-of-Thoughts (2025.acl-long)
Copied to clipboard
| Challenge: | Retrieval-Augmented Generation (RAG) models enable Large Language Models to access external knowledge. |
| Approach: | They propose a knowledge refinement method that incorporates reranking signals to generate CoT-based summarization based on query and retrieval documents. |
| Outcome: | RankCoT generates CoT-based summarization based on query and all retrieval documents . Rank CoT incorporates a self-reflection mechanism that refines the outputs . |
Lost in Literalism: How Supervised Training Shapes Translationese in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models exhibit translationese errors and generate unexpected unnatural translations . Neural machine translation (NMT) has become the dominant method in machine translation research . |
| Approach: | They evaluate the prevalence of translationese in LLM-generated translations and investigate its roots during supervised fine-tuning. |
| Outcome: | The proposed methods reduce translationese while improving translation naturalness . the proposed methods are validated by human evaluations and automatic metrics . |
Accurate KV Cache Quantization with Outlier Tokens Tracing (2025.acl-long)
Copied to clipboard
Yi Su, Yuechi Zhou, Quantong Qiu, Juntao Li, Qingrong Xia, Ping Li, Xinyu Duan, Zhefeng Wang, Min Zhang
| Challenge: | Large Language Models (LLMs) require substantial computational resources during deployment. |
| Approach: | They propose a method to identify outlier tokens and exclude them from quantization . they find that the method can deliver a 6.4 times reduction in memory usage and a 2.5 times increase in throughput . |
| Outcome: | The proposed method delivers a 6.4 times reduction in memory usage and a 2.5 times increase in throughput under 2-bit quantization. |
Can Large Language Models Understand Internet Buzzwords Through User-Generated Content (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for generating definitions of internet buzzwords rely on user-generated content, such as posts and reviews, to understand them. |
| Approach: | They propose a method to generate accurate buzzword definitions using UGC as examples. |
| Outcome: | The proposed method mirrors human language learning skills and can produce more accurate definitions. |
EAC-MoE: Expert-Selection Aware Compressor for Mixture-of-Experts Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Mixture-of-Experts (MoE) has demonstrated promising potential in scaling LLMs . however, it is hindered by two critical challenges: substantial GPU memory consumption and low activated parameters. |
| Approach: | They propose an Expert-Selection Aware Compressor for Mixture-of-Experts (MoE) that aligns with the characteristics of MoE from the perspectives of quantization and pruning. |
| Outcome: | The proposed approach significantly reduces memory usage and improves inference speed with minimal performance degradation. |
Activation Steering Decoding: Mitigating Hallucination in Large Vision-Language Models through Bidirectional Hidden State Intervention (2025.acl-long)
Copied to clipboard
| Challenge: | Large Vision Language Models (LVLMs) suffer from hallucination where generated textual descriptions fail to align accurately with visual semantics. |
| Approach: | They propose a training-free approach that mitigates hallucination through targeted intervention in the model’s intermediate activations by identifying directional patterns of hallucinism in the activation space using a small calibration set. |
| Outcome: | The proposed approach reduces hallucination across multiple benchmarks while maintaining performance on general visual understanding tasks. |
Interactive Evolution: A Neural-Symbolic Self-Training Framework For Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to fine-tune Large Language Models without human annotations are lacking in the field of natural language training. |
| Approach: | They propose an environment-guided neural-symbolic self-training framework to overcome two main challenges: the scarcity of symbolic data and the limited proficiency of LLMs in processing symbolic language. |
| Outcome: | The proposed framework overcomes two main challenges: the scarcity of symbolic data, and the limited proficiency of LLMs in processing symbolic language. |
Improving Medical Large Vision-Language Models with Abnormal-Aware Feedback (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Medical Large Vision-Language Models (Med-LVLMs) lack visual localization in medical images, which is crucial for abnormality detection and interpretation. |
| Approach: | They propose a medical abnormalities unveiling method based on a Medical Abnormalities Unveiler dataset and propose 'abnormal-aware instruction tuning' and 'abbnormal-Aware Reward' method generates diagnoses based upon identified abnormal areas in medical images. |
| Outcome: | The proposed method outperforms existing medical large vision-language models in identifying and understanding medical abnormalities and improves generalization capability. |
Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for tuning large language models from dense to MoE face significant data requirements and require large-scale post-training. |
| Approach: | They propose an upcycling instruction tuning approach for tuning a dense pre-trained model into a MoE instruction model using genetic algorithm and parameter merging. |
| Outcome: | The proposed approach improves the performance of large language models with a small amount of seed data and improves their scaling. |
MapNav: A Novel Memory Representation via Annotated Semantic Maps for VLM-based Vision-and-Language Navigation (2025.acl-long)
Copied to clipboard
Lingfeng Zhang, Xiaoshuai Hao, Qinwen Xu, Qiang Zhang, Xinyao Zhang, Pengwei Wang, Jing Zhang, Zhongyuan Wang, Shanghang Zhang, Renjing Xu
| Challenge: | Vision-language navigation (VLN) is a key task in Embodied AI . traditional approaches rely on historical observations as spatio-temporal contexts for decision making . |
| Approach: | They propose a vision-language navigation model that leverages an annotation system to replace historical frames. |
| Outcome: | The proposed model can be used as a new memory representation method in vision-language navigation . it can be applied to simulated and real-world environments, and it is validated by experiments . |
Exploring Compositional Generalization of Multimodal LLMs for Medical Imaging (2025.acl-long)
Copied to clipboard
Zhenyang Cai, Junying Chen, Rongsheng Wang, Weihong Wang, Yonglin Deng, Dingjie Song, Yize Chen, Zixu Zhang, Benyou Wang
| Challenge: | Current research suggests that multitask training outperforms single-task as different tasks can benefit each other, but they often overlook the internal relationships within these tasks. |
| Approach: | They employ compositional generalization (CG) to examine the generalization of multimodal large language models in medical imaging. |
| Outcome: | The proposed model can understand unseen medical images and is able to perform CG across classification and detection tasks. |
CLAIM: Mitigating Multilingual Object Hallucination in Large Vision-Language Models with Cross-Lingual Attention Intervention (2025.acl-long)
Copied to clipboard
Zekai Ye, Qiming Li, Xiaocheng Feng, Libo Qin, Yichong Huang, Baohang Li, Kui Jiang, Yang Xiang, Zhirui Zhang, Yunfei Lu, Duyu Tang, Dandan Tu, Bing Qin
| Challenge: | Large Vision-Language Models (LVLMs) have impressive multimodal abilities but remain prone to multilingual object hallucination. |
| Approach: | They propose a cross-lingual attention intervention method to mitigate multilingual object hallucination in LVLMs by aligning attention patterns. |
| Outcome: | The proposed method improves 13.56% (up to 30%) on the POPE and 21.75% on the hallucination subsets across languages. |
Wizard of Shopping: Target-Oriented E-commerce Dialogue Generation with Decision Tree Branching (2025.acl-long)
Copied to clipboard
Xiangci Li, Zhiyu Chen, Jason Ingyu Choi, Nikhita Vedula, Besnik Fetahu, Oleg Rokhlenko, Shervin Malmasi
| Challenge: | Prior human-annotated CPS datasets are small in size and lack integration with real-world product search systems. |
| Approach: | They propose a method to generate target-oriented shopping conversations without human annotations by using large language models. |
| Outcome: | The proposed method achieves highly natural and coherent conversations from three shopping domains and significantly improves on human evaluations and downstream tasks. |
Qwen2.5-xCoder: Multi-Agent Collaboration for Multilingual Code Instruction Tuning (2025.acl-long)
Copied to clipboard
Jian Yang, Wei Zhang, Yibo Miao, Shanghaoran Quan, Zhenhe Wu, Qiyao Peng, Liqun Yang, Tianyu Liu, Zeyu Cui, Binyuan Hui, Junyang Lin
| Challenge: | Existing methods to train code LLMs view each programming language in isolation . experimental results show that Qwen2.5-xCoder can bridge the gap between different programming languages . |
| Approach: | They propose a framework that allows agents to collaborate to enhance multilingual instruction tuning for code LLMs. |
| Outcome: | Experimental results show that Qwen2.5-xCoder can transfer knowledge efficiently and effectively between languages. |
Cultivating Gaming Sense for Yourself: Making VLMs Gaming Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Recent efforts leverage Vision Language Models (VLMs) as direct controllers, often pausing the game to analyze screens and plan action through language reasoning. |
| Approach: | They propose a paradigm shift in gameplay agent design that uses Vision Language Models as a developer instead of direct control. |
| Outcome: | The proposed framework achieves fluent gameplay in diverse genres, including ACT, FPS, and Flappy Bird, setting a new benchmark for game-playing agents. |
Genius: A Generalizable and Purely Unsupervised Self-Training Framework For Advanced Reasoning (2025.acl-long)
Copied to clipboard
Fangzhi Xu, Hang Yan, Chang Ma, Haiteng Zhao, Qiushi Sun, Kanzhi Cheng, Junxian He, Jun Liu, Zhiyong Wu
| Challenge: | Existing methods for enhancing LLM reasoning rely on supervisory signals . current methods rely heavily on outcome supervision and auxiliary reward models . |
| Approach: | They propose a gen-eralizable and purely unsupervised self-training framework to enhance LLM reasoning without supervision. |
| Outcome: | The proposed framework improves LLM reasoning without supervision without external supervision. |
Extending Complex Logical Queries on Uncertain Knowledge Graphs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on logical queries on knowledge graphs overlook the incompleteness of KGs. |
| Approach: | They propose an ML-based approach to answer soft queries on uncertain knowledge . they propose to use forward inference and backward calibration to avoid catastrophic errors . |
| Outcome: | The proposed method ensures there are no catastrophic cascading errors while maintaining the same complexity as state-of-the-art inference algorithms for first-order queries. |
Knowledge Decoupling via Orthogonal Projection for Lifelong Editing of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for enhancing large language models (LLMs) have achieved some success, but their knowledge understanding and memory capacity significantly degrades after extensive editing. |
| Approach: | They propose a method that stores the basis vectors of the representation space of past edits in a knowledge cache and projects the gradient of the current edit onto a space orthogonal to previous knowledge for updating. |
| Outcome: | The proposed method improves question-answering ability and hallucination mitigation by 14% and 61% for large language models after 3,000 edits. |
𝜙-Decoding: Adaptive Foresight Sampling for Balanced Inference-Time Exploration and Exploitation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing inference-time optimization strategies address the shortsightedness of auto-regressive generation, but the vast search space leads to excessive exploration and insufficient exploitation. |
| Approach: | They propose a decoding strategy that approximates two distributions via foresight and clustering to provide an efficient estimation of step value. |
| Outcome: | The proposed decoding strategy outperforms strong baselines in performance and efficiency. |
Can LLM Watermarks Robustly Prevent Unauthorized Knowledge Distillation? (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Model (LLM) watermarking is radioactive and enables the detection of watermarks inherited by student models when trained on the outputs of watermarked teacher models. |
| Approach: | They propose two types of watermark removal attacks that allow student models to perform untraceable knowledge distillation while avoiding watermark inheritance. |
| Outcome: | The proposed attacks eliminate inherited watermarks while maintaining knowledge transfer efficiency and low computational overhead. |
Rethinking Reward Model Evaluation Through the Lens of Reward Overoptimization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks for reward models show a weak correlation with performance of optimized policies . existing benchmarks do not accurately assess the true capabilities of reward models . |
| Approach: | They explore how reward overoptimization captures how well a reward model aligns with human preferences and the dynamics of the learning signal it provides to the policy. |
| Outcome: | The proposed benchmarks show that reward overoptimization is a weak factor . the high correlation with degree of overoptimalization leads to lower correlation with downstream performance . |
Inducing lexicons of in-group language with socio-temporal context (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for lexicon induction do not capture the evolving nature of in-group language, nor the social structure of the community. |
| Approach: | They propose a method for inducing lexicons of in-group language which incorporates its socio-temporal context. |
| Outcome: | The proposed method outperforms existing methods for lexicon induction . it quantifies relevance of each term to a specific sub-community at a given point in time . |
LLaSE-G1: Incentivizing Generalization Capability for LLaMA-based Speech Enhancement (2025.acl-long)
Copied to clipboard
Boyi Kang, Xinfa Zhu, Zihan Zhang, Zhen Ye, Mingshuai Liu, Ziqian Wang, Yike Zhu, Guobin Ma, Jun Chen, Longshuai Xiao, Chao Weng, Wei Xue, Lei Xie
| Challenge: | Recent advances in language models have demonstrated strong capabilities in semantic understanding and contextual modeling. |
| Approach: | They propose a LLaMA-based language model that incentivizes generalization capabilities for speech enhancement. |
| Outcome: | The proposed language model outperforms prior task-specific discriminative and generative models in acoustic enhancement tasks. |
MadaKV: Adaptive Modality-Perception KV Cache Eviction for Efficient Multimodal Long-Context Inference (2025.acl-long)
Copied to clipboard
Kunxi Li, Zhonghua Jiang, Zhouzhou Shen, ZhaodeWang ZhaodeWang, Chengfei Lv, Shengyu Zhang, Fan Wu, Fei Wu
| Challenge: | Existing KV cache eviction methods fail to capture modality-specific information, resulting in suboptimal performance. |
| Approach: | They propose a modality-adaptive key-value (KV) cache eviction strategy to enhance the efficiency of multimodal large language models in long-context inference. |
| Outcome: | The proposed method reduces the KV cache memory footprint and model inference latency while maintaining high accuracy across multimodal long-context tasks. |
Efficient OpAmp Adaptation for Zoom Attention to Golden Contexts (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models have shown significant promise in question-answering tasks . noisy reference documents hinder performance of LLMs, causing disproportionate attention to irrelevant content . |
| Approach: | They propose an adaptive large language model that allocates disproportionate attention to irrelevant documents . they use transformers to train the model and integrate it into pre-trained Transformer blocks . |
| Outcome: | The proposed model outperforms state-of-the-art models on noisy-context benchmarks. |
Language-Codec: Bridging Discrete Codec Representations and Speech Language Models (2025.acl-long)
Copied to clipboard
Shengpeng Ji, Minghui Fang, Jialong Zuo, Ziyue Jiang, Dingdong Wang, Hanting Wang, Hai Huang, Zhou Zhao
| Challenge: | Existing gaps between discrete acoustic codecs and downstream speech language models . initial channel of codebooks contains excessive information, making it difficult to generate tokens from weakly supervised signals such as text. |
| Approach: | They propose a discrete acoustic codec for generating acustic tokens from weakly supervised signals. |
| Outcome: | The proposed language-codec outperforms competing audio compression algorithms and validates on downstream speech language models. |
Adaptive Tool Use in Large Language Models with Meta-Cognition Trigger (2025.acl-long)
Copied to clipboard
Wenjun Li, Dexun Li, Kuicai Dong, Cong Zhang, Hao Zhang, Weiwen Liu, Yasheng Wang, Ruiming Tang, Yong Liu
| Challenge: | Existing research expands the tool arrays of large language models (LLMs), but the necessity of using these tools is often overlooked, leading to indiscriminate tool invocation. |
| Approach: | They propose a meta-cognition proxy proxy for LLMs self-assessment of their capabilities, reflecting the model’s awareness of its own limitations. |
| Outcome: | The proposed strategy is fine-tuned-free and costs minimal. |
MMLU-CF: A Contamination-free Multi-task Language Understanding Benchmark (2025.acl-long)
Copied to clipboard
Qihao Zhao, Yangyu Huang, Tengchao Lv, Lei Cui, Qinzheng Sun, Shaoguang Mao, Xin Zhang, Ying Xin, Qiufeng Yin, Scarlett Li, Furu Wei
| Challenge: | Multiple-choice question datasets like Massive Multitask Language Understanding (MMLU) have inevitably led to benchmark contamination, resulting in unreliable evaluation. |
| Approach: | They propose a contamination-free MCQ benchmark called MMLU-CF which reassesses LLMs’ understanding of world knowledge by averting both unintentional and malicious data contamination. |
| Outcome: | The proposed MMLU-CF reassesses LLMs’ understanding of world knowledge by averting both unintentional and malicious data contamination. |
Code-Switching Red-Teaming: LLM Evaluation for Safety and Multilingual Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Recent large language models (LLMs) are inherently multilingual agents . concerns regarding their safety have emerged . |
| Approach: | They propose a framework to synthesize red-teaming queries and investigate their safety . they demonstrate that the framework outperforms existing red- teaming techniques . |
| Outcome: | The proposed framework outperforms existing red-teaming techniques in the safety domain . it generates code-switching attack prompts in monolingual data . |
Unleashing LLM Reasoning Capability via Scalable Question Synthesis from Scratch (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve the mathematical reasoning capabilities of Large Language Models (LLMs) are limited due to the proprietary nature of the data. |
| Approach: | They propose a data synthesis method that generates large-scale mathematical reasoning datasets using lightweight 7B-scale models. |
| Outcome: | The proposed method outperforms existing open-source datasets in both in-domain and out-of-domain evaluations and shows improvements in code reasoning tasks. |
DREsS: Dataset for Rubric-based Essay Scoring on EFL Writing (2025.acl-long)
Copied to clipboard
| Challenge: | Automated essay scoring (AES) is a useful tool in English as a foreign language (EFL) writing education. |
| Approach: | They propose a large-scale, standard dataset for rubric-based automated essay scoring with 48.9K samples in total. |
| Outcome: | The proposed system improves the baseline scores by 45.44%. |
PQR: Improving Dense Retrieval via Potential Query Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing training data is sparse, with each document associated with one or a few labeled queries. |
| Approach: | They propose a training-free potential query retrieval framework to address this problem . they use a Gaussian mixture distribution to model all potential queries for a document . |
| Outcome: | The proposed method is able to capture comprehensive semantic information from a document with multiple queries. |
Cross-Lingual Generalization and Compression: From Language-Specific to Shared Neurons (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evidence suggests that multilingual language models can transfer knowledge across languages without explicit cross-lingual supervision. |
| Approach: | They analyze the parameter spaces of three multilingual language models to examine their representations . they find that models evolve from language-specific representations to more specialized layer functions . |
| Outcome: | The proposed model can generate coherent English text, rather than spanish text, and it can generate generalized representations, the authors show. |
SDBench: A Survey-based Domain-specific LLM Benchmarking and Optimization Framework (2025.acl-long)
Copied to clipboard
| Challenge: | acquiring domain-specific knowledge often requires professional expert manpower. |
| Approach: | They propose a generic framework for generating evaluation datasets for domain-specific LLMs. |
| Outcome: | The proposed framework reduces the reliance on expert manpower while ensuring that the collected data is uniformly distributed. |
ReflecTool: Towards Reflection-Aware Tool-Augmented Clinical Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown promising potential in the medical domain, assisting with tasks like clinical note generation and patient communication. |
| Approach: | They propose a framework that excels at utilizing domain-specific tools within two stages. |
| Outcome: | The proposed framework surpasses the pure LLMs with more than 10 points and the well-established agent-based methods with 3 points. |
Lexical Recall or Logical Reasoning: Probing the Limits of Reasoning Abilities in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on LLMs assesses logic abilities independently from lexical memory. |
| Approach: | They propose to assess LLMs' logic abilities independently from lexical memory . they use two sets of grid puzzles in 42 different sizes and 12 difficulty levels . |
| Outcome: | The proposed benchmarks show that LLMs are limited to a few steps of reasoning . the results show that the applied obfuscation strategies help mitigate effects of logic puzzles being part of training data. |
ChainEdit: Propagating Ripple Effects in LLM Knowledge Editing through Logical Rule-Guided Chains (2025.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge editing methods for large language models struggle to maintain logical consistency when propagating ripple effects to associated facts. |
| Approach: | They propose a framework that synergizes knowledge graph-derived logical rules with LLM logical reasoning capabilities to enable systematic chain updates. |
| Outcome: | The proposed framework improves logical generalization and specificity while maintaining reliability and specificness. |
HiDe-LLaVA: Hierarchical Decoupling for Continual Instruction Tuning of Multimodal Large Language Model (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve instructionfollowing performance of MLLMs often trade off memory efficiency for performance gains, compromising overall efficiency. |
| Approach: | They propose a task-specific expansion and task-general fusion framework based on variations in Centered Kernel Alignment (CKA) similarity across different model layers when trained on diverse datasets. |
| Outcome: | The proposed framework improves performance compared to existing benchmarks. |
Self-supervised Quantized Representation for Seamlessly Integrating Knowledge Graphs with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are gaining popularity due to their lack of knowledge hallucination and lack of a coherent model. |
| Approach: | They propose a self-supervised quantized representation method to compress KG structural and semantic knowledge into discrete codes that align the format of language sentences. |
| Outcome: | The proposed framework outperforms existing unsupervised methods producing more distinguishable codes on KG link prediction and triple classification tasks. |
Finite State Automata Inside Transformers with Chain-of-Thought: A Mechanistic Study on State Tracking (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that Chain-of-thought (CoT) can enhance the performance of large language models (LLMs) however, there is limited understanding of the algorithms that Transformer+CoT can learn. |
| Approach: | They propose two metrics to evaluate Transformer+CoT's state tracking capabilities and identify the circuit responsible for tracking the world state. |
| Outcome: | The proposed model achieves 100% accuracy for each state, highlighting an implicit finite state automaton (FSA) embedded within the model. |
TeamLoRA: Boosting Low-Rank Adaptation with Expert Collaboration and Competition (2025.acl-long)
Copied to clipboard
Tianwei Lin, Jiang Liu, Wenqiao Zhang, Yang Dai, Haoyuan Li, Zhelun Yu, Wanggui He, Juncheng Li, Jiannan Guo, Hao Jiang, Siliang Tang, Yueting Zhuang
| Challenge: | Existing methods for fine-tuning are resource-efficient, but performance often falls short . a new approach, TeamLoRA, integrates collaborative and competitive modules to improve performance. |
| Approach: | They propose to introduce task-specific LoRA as domain experts to improve learning efficiency . teamLoRA integrates collaborative and competition modules to improve model learning . |
| Outcome: | Experiments show that TeamLoRA improves performance in multi-task learning . teamLorea integrates collaborative and competitive modules to improve performance . |
CRiskEval: A Chinese Multi-Level Risk Evaluation Benchmark Dataset for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models exhibit a risk propensity of more than 40%, and are prone to self-sustainability, power seeking and other dangerous goals as the size of models increases. |
| Approach: | They propose a Chinese dataset to gauge the risk proclivities inherent in large language models such as resource acquisition and malicious coordination. |
| Outcome: | The proposed dataset consists of 14,888 questions that simulate scenarios related to predefined 7 types of frontier risks. |
STUN: Structured-Then-Unstructured Pruning for Scalable MoE Pruning (2025.acl-long)
Copied to clipboard
| Challenge: | Mixture-of-experts (MoEs) have been adopted for reducing inference costs by sparsely activating experts in large language models (LLMs). |
| Approach: | They propose a structured-then-unstructured approach outperforming both of structured and unstructured pruning for MoEs. |
| Outcome: | The proposed approach outperforms both of structured and unstructured pruning, especially for MoEs with hundreds of experts. |
Mimicking the Familiar: Dynamic Command Generation for Information Theft Attacks in LLM Tool-Learning System (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to attack Large Language Model (LLM) tool-learning systems are black-box oriented and rely on static commands that cannot adapt flexibly to the changes in user queries and the invocation chain. |
| Approach: | They propose a dynamic attack comment generation approach for information theft attacks in LLM tool-learning systems that mimics the familiar by inferring the information utilized by upstream tools. |
| Outcome: | The proposed approach outperforms baselines with +13.2% ASRTheft and can be generalized to new tool-learning systems to expose their information leakage risks. |
FlashAudio: Rectified Flow for Fast and High-Fidelity Text-to-Audio Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in latent diffusion models (LDMs) have markedly enhanced text-to-audio generation, yet their iterative sampling processes impose substantial computational demands, limiting practical deployment. |
| Approach: | They propose to learn straight flow for fast simulation by using flashAudio with rectified flows and immiscible flow to minimize the total distance of data-noise pairs in a batch vias assignment. |
| Outcome: | The proposed method can learn straight flow for fast simulations and reduce noise distribution. |
How does Misinformation Affect Large Language Model Behaviors and Preferences? (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have explored the role of Large Language Models in combating misinformation, but there is still a lack of detailed analysis on the specific aspects and extent to which LLMs are influenced by misinformation. |
| Approach: | They propose to use a benchmark to evaluate LLMs' behavior and knowledge preference toward misinformation to identify their models. |
| Outcome: | The proposed approach is based on 10,346,712 pieces of misinformation and examines knowledge conflicts and stylistic variations. |
YESciEval: Robust LLM-as-a-Judge for Scientific Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) drive scientific question-answering on search engines, yet their evaluation robustness remains underexplored. |
| Approach: | They propose an open-source framework that combines rubric-based assessment with reinforcement learning to mitigate optimism bias in LLM evaluators. |
| Outcome: | The proposed framework combines fine-grained rubric-based assessment with reinforcement learning to mitigate optimism bias in LLM evaluators. |
GALLa: Graph Aligned Large Language Models for Improved Source Code Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Programming languages have rich semantics that are represented by graphs and not available from the surface form of source code. |
| Approach: | They propose to use graph neural networks and cross-modal alignment technologies to inject structural information of code into LLMs as an auxiliary task during finetuning. |
| Outcome: | The proposed framework improves on five code tasks with six different baseline LLMs, while incurring no cost at inference time. |
MEDDxAgent: A Unified Modular Agent Framework for Explainable Automatic Differential Diagnosis (2025.acl-long)
Copied to clipboard
Daniel Philip Rose, Chia-Chien Hung, Marco Lepri, Israa Alqassem, Kiril Gashteovski, Carolin Lawrence
| Challenge: | Existing approaches to DDx are limited by single-dataset evaluations, isolated optimization of components, unrealistic assumptions about complete patient profiles, and single-attempt diagnosis. |
| Approach: | They propose a Modular Explainable DDx Agent framework that allows physicians to iteratively refine a ranked list of possible diseases based on symptoms, antecedents, and medical knowledge. |
| Outcome: | The proposed framework achieves over 10% accuracy improvements in interactive DDx across large and small LLMs while offering critical explainability into its diagnostic reasoning process. |
A Training-free LLM-based Approach to General Chinese Character Error Correction (2025.acl-long)
Copied to clipboard
| Challenge: | Chinese spelling correction (CSC) is a crucial task that aims to correct character errors in text. |
| Approach: | They propose a task that handles missing and redundant characters and an additional prompt-based large language model to improve performance. |
| Outcome: | The proposed task is based on a high-quality dataset and a prompt-based large language model. |
HSCR: Hierarchical Self-Contrastive Rewarding for Aligning Medical Vision Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for medical vision-language models overlook modality misalignment . HSCR generates high-quality preference data with higher sampling probability . |
| Approach: | They propose a hierarchical self-contrastive reward approach that addresses two challenges in alignment . they leverage the inherent capability of Med-VLMs to generate dispreferred responses . |
| Outcome: | The proposed approach improves accuracy and trustworthiness of medical vision-label models with 2,000 training entries. |
MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale (2025.acl-long)
Copied to clipboard
Jiawei Guo, Tianyu Zheng, Yizhi Li, Yuelin Bai, Bo Li, Yubo Wang, King Zhu, Graham Neubig, Wenhu Chen, Xiang Yue
| Challenge: | Current instruction-tuning datasets focus on simplistic visual question answering tasks, and provide phrase-level answers without any intermediate rationales. |
| Approach: | They propose to use open-source multimodal large language models to train MLLMs on a dataset with 12M instruction-response pairs to elicit CoT reasoning. |
| Outcome: | The proposed model achieves state-of-the-art performance on benchmarks such as MathVerse, MMMU-Pro, and MuirBench, and gains improvements of up to 4% on non-reasoning-based benchmarks. |
SIFT-50M: A Large-Scale Multilingual Dataset for Speech Instruction Fine-Tuning (2025.acl-long)
Copied to clipboard
Prabhat Pandey, Rupak Vignesh Swaminathan, K V Vijay Girish, Arunasish Sen, Jian. Xie, Grant Strimel, Andreas Schwarz
| Challenge: | Recent years have witnessed significant advancements in integrating speech and audio capabilities into large language models. |
| Approach: | They propose a 50M-example dataset for instruction fine-tuning and pre-training of speech-text large language models (LLMs) the dataset spans five languages and enables a diverse range of speech understanding and controllable speech generation instructions. |
| Outcome: | The proposed dataset outperforms existing speech-text LLMs on instruction-following benchmarks while achieving competitive performance on foundational speech tasks. |
Recent Advances in Speech Language Models: A Survey (2025.acl-long)
Copied to clipboard
Wenqian Cui, Dianzhi Yu, Xiaoqi Jiao, Ziqiao Meng, Guangyan Zhang, Qichao Wang, Steven Y. Guo, Irwin King
| Challenge: | Text-based Large Language Models (LLMs) are a promising solution for end-to-end speech interaction. |
| Approach: | They propose to build a framework that allows users to input text and translate it into speech . they propose to use a text-only LLM and a "textto-speech" framework to generate a response based on this transcription . |
| Outcome: | The survey offers an overview of recent approaches to building SpeechLMs . it outlines core architectural components, training methodologies, evaluation strategies and challenges . |
LexCLiPR: Cross-Lingual Paragraph Retrieval from Legal Judgments (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on IR focus on retrieving entire cases rather than precise, paragraph-level information. |
| Approach: | They propose a cross-lingual dataset for paragraph-level retrieval from ECtHR judgments . they evaluate retrieval models in a zero-shot setting and use multilingual case law guides . |
| Outcome: | The proposed model excels in cross-lingual retrieval, while siamese architectures are better suited for monolingual tasks. |
Multi-task Adversarial Attacks against Black-box Model with Few-shot Queries (2025.acl-long)
Copied to clipboard
| Challenge: | Existing adversarial text attacks rely on abundant access to shared internal features and numerous queries, limited to a single task type. |
| Approach: | They propose a black-box attack that exploits the transferability of adversarial texts . they use a deep-level substitute model trained in a plug-and-play manner for text classification . |
| Outcome: | The proposed attack can target multiple tasks with minimal perturbations . it can target commercial APIs, large language models, and image-generation models . |
SPECTRA: Faster Large Language Model Inference with Optimized Internal and External Speculation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to inference with Large Language Models (LLMs) are expensive and time-consuming. |
| Approach: | They propose a framework for accelerating large language model inference without additional training or modification to the original LLM. |
| Outcome: | The proposed framework outperforms state-of-the-art methods and achieves 4.08x speedups across benchmarks and LLM architectures. |
Multi-level Association Refinement Network for Dialogue Aspect-based Sentiment Quadruple Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for identifying quadruples rely on predefined dialogue structure and word semantics to achieve accurate and comprehensive sentiment associations between utterances and words. |
| Approach: | They propose a multi-level association refinement network to achieve more accurate sentiment associations between utterances and words. |
| Outcome: | The proposed framework achieves state-of-the-art performance under low-resource conditions. |
Innovative Image Fraud Detection with Cross-Sample Anomaly Analysis: The Power of LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for document image fraud detection lack visual clues on tampered regions. |
| Approach: | They propose a framework for detecting logical inconsistencies in document images by leveraging LLMs. |
| Outcome: | The proposed framework outperforms state-of-the-art fraud detection methods by 79.6% on CrossCred and industrial solutions by 21.7% on business data. |
Cooperative or Competitive? Understanding the Interaction between Attention Heads From A Game Theory Perspective (2025.acl-long)
Copied to clipboard
| Challenge: | a number of attention-based large language models (LLMs) focus on individual head contributions, but the precise interaction mechanisms between attention heads remain poorly understood. |
| Approach: | They propose a game-theoretic attention calibration method that uses the Harsanyi dividend . they selectively retain heads demonstrating significant cooperative gains and apply fine-grained adjustments to remaining heads . |
| Outcome: | The proposed framework is based on the Harsanyi dividend, a concept from cooperative game theory. |
MM-Verify: Enhancing Multimodal Reasoning with Chain-of-Thought Verification (2025.acl-long)
Copied to clipboard
| Challenge: | MM-Verifier and MM Reasoner are a powerful multimodal reasoning model . large language models (LLMs) have demonstrated exceptional performance across tasks spanning myriad domains. |
| Approach: | They propose a method which combines tree search and verification to generate high-quality chain-of-thought data. |
| Outcome: | The proposed method outperforms all larger models on the MathCheck, MathVista, and MathVerse benchmarks. |
Graph-Structured Trajectory Extraction from Travelogues (2025.acl-long)
Copied to clipboard
Aitaro Yamamoto, Hiroyuki Otomo, Hiroki Ouchi, Shohei Higashiyama, Hiroki Teranishi, Hiroyuki Shindo, Taro Watanabe
| Challenge: | Existing studies treat travelogues as sequences of visited locations, but they lack a benchmark dataset. |
| Approach: | They propose to represent the trajectory as a graph that can capture the hierarchy as well as the visiting order and construct a benchmark dataset for the extraction. |
| Outcome: | The proposed dataset shows that even naive baseline systems can predict visited locations and the visiting order between them, while it is more challenging to predict the hierarchical relations. |
Learning First-Order Logic Rules for Argumentation Mining (2025.acl-long)
Copied to clipboard
Yang Sun, Guanrong Chen, Hamid Alinejad-Rokny, Jianzhu Bao, Yuqi Huang, Bin Liang, Kam-Fai Wong, Min Yang, Ruifeng Xu
| Challenge: | Argumentation Mining (AM) aims to extract argumentative structures from texts by identifying argumentation components (ACs) and their argumentative relations (ARs). |
| Approach: | They propose a First- Order Logic reasoning framework for AM to capture logical reasoning paths within argumentative texts. |
| Outcome: | The proposed framework outperforms strong baselines while significantly improving explainability. |
Investigating and Enhancing the Robustness of Large Multimodal Models Against Temporal Inconsistency (2025.acl-long)
Copied to clipboard
Jiafeng Liang, Shixin Jiang, Xuan Dong, Ning Wang, Zheng Chu, Hui Su, Jinlan Fu, Ming Liu, See-Kiong Ng, Bing Qin
| Challenge: | Large Multimodal Models (LMMs) have demonstrated impressive performance on general video comprehension benchmarks, but their robustness needs to be thoroughly investigated for broader applications. |
| Approach: | They propose a temporal robustness benchmark which introduces temporal inconsistency perturbations separately at the visual and textual modalities to assess the robustness of models. |
| Outcome: | The proposed method improves the model’s robustness and reliability in temporal analysis. |
UniRAG: Unified Query Understanding Method for Retrieval Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing query augmentation methods face knowledge update lag and hallucinations in large language models (LLMs) Existing methods face two key challenges: (1) separation of query augmented and encoding tasks, which hinders information sharing and introduces cumulative errors; (2) difficulty of selecting optimal augmentation strategy for different scenarios. |
| Approach: | They propose a unified framework for query understanding in RAG that integrates internal and external knowledge to enhance query augmentation and encoding tasks. |
| Outcome: | The proposed framework outperforms traditional query augmentation methods in five knowledge-intensive benchmark tasks in both closed and open domain question answering. |
Contextual Experience Replay for Self-Improvement of Language Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Existing large language model (LLM) agents fail in complex tasks without any environment-specific experiences. |
| Approach: | They propose a framework that accumulates and synthesizes past experiences into a dynamic memory buffer to enable efficient self-improvement for language agents in their context window. |
| Outcome: | The proposed framework improves performance on WebArena and VisualWebAren . it surpasses tree search method with fewer token costs and achieves state-of-the-art performance of 31.9%. |
Emma-X: An Embodied Multimodal Action Model with Grounded Chain of Thought and Look-ahead Spatial Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Visual-Language-Action models lack the ability to generate actionable policies tailored to specific robotic embodiments. |
| Approach: | They propose an embodied multimodal action model with Grounded Chain of Thought and Look-ahead Spatial Reasoning that enhances spatial reasoning and task planning. |
| Outcome: | The proposed model improves on existing baselines in tasks requiring spatial reasoning and grounding reasoning. |
Towards Comprehensive Argument Analysis in Education: Dataset, Tasks, and Method (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research on argument mining has proposed various argument annotation schemes and tasks. |
| Approach: | They propose a framework comprising 14 fine-grained relation types to capture the interplay between argument components for a thorough understanding of argument structure. |
| Outcome: | The proposed framework captures the interplay between argument components for a thorough understanding of argument structure. |
Browsing Like Human: A Multimodal Web Agent with Experiential Fast-and-Slow Thinking (2025.acl-long)
Copied to clipboard
| Challenge: | Existing web agents lack visual perception, planning, and memory abilities, but their reasoning process is deviate from human cognition. |
| Approach: | They propose a multimodal web agent framework that emulates human planning process to decompose complex user instructions. |
| Outcome: | The proposed framework emulates human planning process to decompose complex user instructions. |
MaXIFE: Multilingual and Cross-lingual Instruction Following Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods focus on single-language scenarios, overlooking multilingual and cross-lingual contexts. |
| Approach: | They propose a tool to assess instruction-following capabilities across 23 different languages with 1667 verifiable instruction tasks. |
| Outcome: | MaXIFE evaluates instruction-following capabilities across 23 languages with 1667 verifiable instruction tasks. |
Linguistic Generalizability of Test-Time Scaling in Mathematical Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies show that pre-training compute can improve multilingual performance, but is it effective for test-time scaling? |
| Approach: | They propose a multilingual math benchmark with competition-level problems in 55 languages . they propose "test-time scaling" which further lengthens the time it takes to scale . |
| Outcome: | The proposed methods fail to generalize robustly across languages, with no improvements in variance or consistency. |
Can MLLMs Understand the Deep Implication Behind Chinese Images? (2025.acl-long)
Copied to clipboard
Chenhao Zhang, Xi Feng, Yuelin Bai, Xeron Du, Jinchang Hou, Kaixin Deng, Guangzeng Han, Qinrui Li, Bingli Wang, Jiaheng Liu, Xingwei Qu, Yifei Zhang, Qixuan Zhao, Yiming Liang, Ziqiang Liu, Feiteng Fang, Min Yang, Wenhao Huang, Chenghua Lin, Ge Zhang, Shiwen Ni
| Challenge: | MLLMs perform poorly on traditional culture images, indicating limitations in understanding high-level semantics and lacking a deep knowledge base of Chinese traditional culture. |
| Approach: | They propose to use Chinese images to assess MLLMs' higher-order perception and understanding of Chinese visual content. |
| Outcome: | The proposed model incorporates images that represent Chinese traditional culture, such as famous Chinese traditional paintings, to ensure the authenticity of the Chinese context. |
KazMMLU: Evaluating Language Models on Kazakh, Russian, and Regional Knowledge of Kazakhstan (2025.acl-long)
Copied to clipboard
Mukhammed Togmanov, Nurdaulet Mukhituly, Diana Turmakhan, Jonibek Mansurov, Maiya Goloburda, Akhmed Sakip, Zhuohan Xie, Yuxia Wang, Bekassyl Syzdykov, Nurkhan Laiyk, Alham Fikri Aji, Ekaterina Kochmar, Preslav Nakov, Fajri Koto
| Challenge: | Kazakh language remains underrepresented in the field of natural language processing despite the country's population exceeding twenty million . however, there is a lack of dedicated models and benchmark evaluations specifically tailored to Kazakh languages. |
| Approach: | They propose to create a dataset specifically designed for Kazakh language with 23,000 questions sourced from authentic educational materials and manually validated by native speakers and educators. |
| Outcome: | The first MMLU-style dataset specifically designed for Kazakh language. |
Towards Multi-dimensional Evaluation of LLM Summarization across Domains and Languages (2025.acl-long)
Copied to clipboard
Hyangsuk Min, Yuho Lee, Minjeong Ban, Jiaqi Deng, Nicole Hee-Yeon Kim, Taewon Yun, Hang Su, Jason Cai, Hwanjun Song
| Challenge: | Existing evaluation frameworks for text summarization lack domain-specific assessment criteria and are predominantly English-centric. |
| Approach: | They propose a multi-dimensional, multi-domain evaluation of summarization in English and Chinese that incorporates specialized assessment criteria for each domain and leverages a debate system to enhance annotation quality. |
| Outcome: | The proposed evaluation framework provides a multi-dimensional, multi-domain evaluation of summarization in English and Chinese. |
ClusterAttn: KV Cache Compression under Intrinsic Attention Clustering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for sparse attention apply the same pattern across different attention heads and inputs, but fail to capture the intrinsic attention clustering in large language models. |
| Approach: | They propose a training-free sparse attention method that provides an efficient prompt cache compression scheme under intrinsic attention clustering for efficient LLM inference. |
| Outcome: | The proposed method reduces memory usage by 10%–65% and increases throughput by 2.6–4.8 times with no accuracy loss. |
SHARE: Shared Memory-Aware Open-Domain Long-Term Dialogue Dataset Constructed from Movie Script (2025.acl-long)
Copied to clipboard
| Challenge: | Antoine de Saint-Exupéry Memory in dialogue plays a crucial role in building relationships and facilitating the ongoing conversation. |
| Approach: | They propose a long-term dialogue dataset named SHARE that includes shared memories between two individuals. |
| Outcome: | The proposed dataset makes long-term dialogues more engaging and sustainable . it includes summaries of persona information and events of two individuals . |
Incongruity-aware Tension Field Network for Multi-modal Sarcasm Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-modal sarcasm detection (MSD) identifies sarcasm and accurately understands users’ real attitudes from text-image pairs. |
| Approach: | They propose to use incongruity-aware tension field network to extract effective text-image feature pairs in fact and sentiment perspectives and construct a fact/sentiment tension field with discrepancy metrics to capture contextual tone and polarized inconcongruities. |
| Outcome: | The proposed method achieves state-of-the-art performance surpassing LLaVA1.5-7B with only 17.3M trainable parameters, demonstrating its optimal performance-efficiency in multi-modal sarcasm detection tasks. |
Instruction Tuning on Public Government and Cultural Data for Low-Resource Language: a Case Study in Kazakh (2025.acl-long)
Copied to clipboard
| Challenge: | Instruction tuning in low-resource languages remains underexplored due to limited text data, particularly in government and cultural domains. |
| Approach: | They propose to open-source a large-scale instruction-following dataset covering key institutional and cultural knowledge relevant to Kazakhstan. |
| Outcome: | The proposed dataset improves LLMs’ understanding of procedural, legal, and structural governance topics. |
Stealing Training Data from Large Language Models in Decentralized Training through Activation Inversion Attack (2025.acl-long)
Copied to clipboard
| Challenge: | Decentralized training is a resource-efficient framework to democratize training of large language models. |
| Approach: | They propose an activation inversion attack to exploit privacy leakage from training data . they construct a shadow dataset comprising text labels and corresponding activations . |
| Outcome: | The proposed attack surface is based on a shadow dataset and public datasets . the proposed attack model reconstructs training data from activations in victim decentralized training. |
From Selection to Generation: A Survey of LLM-based Active Learning (2025.acl-long)
Copied to clipboard
Yu Xia, Subhojyoti Mukherjee, Zhouhang Xie, Junda Wu, Xintong Li, Ryan Aponte, Hanjia Lyu, Joe Barrow, Hongjie Chen, Franck Dernoncourt, Branislav Kveton, Tong Yu, Ruiyi Zhang, Jiuxiang Gu, Nesreen K. Ahmed, Yu Wang, Xiang Chen, Hanieh Deilamsalehy, Sungchul Kim, Zhengmian Hu, Yue Zhao, Nedim Lipka, Seunghyun Yoon, Ting-Hao Kenneth Huang, Zichao Wang, Puneet Mathur, Soumyabrata Pal, Koyel Mukherjee, Zhehao Zhang, Namyong Park, Thien Huu Nguyen, Jiebo Luo, Ryan A. Rossi, Julian McAuley
| Challenge: | Large Language Models (LLMs) have been used for selection and training of data for active learning. |
| Approach: | They propose an intuitive taxonomy that categorizes LLM-based active learning techniques and discuss the transformative roles they can play in the active learning loop. |
| Outcome: | The proposed model can generate entirely new data instances and provide more cost-effective annotations with fewer labeled data instances. |
OmniFlatten: An End-to-end GPT Model for Seamless Voice Conversation (2025.acl-long)
Copied to clipboard
Qinglin Zhang, Luyao Cheng, Chong Deng, Qian Chen, Wen Wang, Siqi Zheng, Jiaqing Liu, Hai Yu, Chao-Hong Tan, Zhihao Du, ShiLiang Zhang
| Challenge: | Full-duplex spoken dialogue systems allow simultaneous bidirectional communication . low latency and natural interactions in full-duplice systems remains a challenge . |
| Approach: | They propose a multi-stage post-training scheme that adapts a text large language model into a speech-text dialogue LLM. |
| Outcome: | The proposed model can model human conversation behaviors with low latency and natural interactions with low delay. |
DoMIX: An Efficient Framework for Exploiting Domain Knowledge in Fine-Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for domain-adaptive pre-training (DAP) face several limitations: high computational cost and GPU memory usage during training; and lack of generalized model for all end tasks. |
| Approach: | They propose a domain-adaptive pre-training (DAP) method that uses a representative parameter-efficient fine-tuning method to provide pre-trained models for specific tasks. |
| Outcome: | The proposed method can be extended beyond the DAP setting to standard LLM fine-tuning scenarios. |
EAGLE: Expert-Guided Self-Enhancement for Preference Alignment in Pathology Large Vision-Language Model (2025.acl-long)
Copied to clipboard
Meidan Ding, Jipeng Zhang, Wenxuan Wang, Haiqin Zhong, Xiaoqin Wang, Xinheng Lyu, Wenting Chen, Linlin Shen
| Challenge: | Recent advances in Large Vision Language Models (LVLMs) show promise for pathological diagnosis, yet their application in clinical settings faces critical challenges of multimodal hallucination and biased responses. |
| Approach: | They propose a framework that integrates medical expertise into preference alignment. |
| Outcome: | The proposed framework outperforms existing pathological LVLMs while maintaining pathological accuracy. |
CoT-ICL Lab: A Synthetic Framework for Studying Chain-of-Thought Learning from In-Context Demonstrations (2025.acl-long)
Copied to clipboard
| Challenge: | In-context learning and CoT are still poorly understood, but the precise mechanisms and architectural factors driving ICL and Co T are still unclear. |
| Approach: | They propose a framework and methodology to generate synthetic tokenized datasets and study chain-of-thought (CoT) in-context learning in language models. |
| Outcome: | The proposed framework and methodology allows fine grained control over the complexity of in-context examples by decoupling causal structure from underlying token processing functions. |
Flexora: Flexible Low-Rank Adaptation for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have revolutionized artificial intelligence, but performance on specific tasks is limited by knowledge boundaries. |
| Approach: | They propose a method that automatically selects the most critical layers for fine-tuning to optimize performance across diverse downstream tasks. |
| Outcome: | The proposed method outperforms baseline models and natural language tasks. |
QDTSynth: Quality-Driven Formal Theorem Synthesis for Enhancing Proving Performance of LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing formal languages such as Lean, Coq and Metamath are proving to be useful in formal theorem proving . however, there is a scarcity of high-quality supervised fine-tuning data for formal proofs . |
| Approach: | They propose a Q**uality-**D**riven **T**heorem **S**ynthesis method in Lean4 . they propose diversity screening and the self-assessment method to select theoremas that exhibit diversity and high quality from the initial synthetic statements. |
| Outcome: | The proposed method significantly improves performance of open-source LLMs in theorem proving tasks. |
RSVP: Reasoning Segmentation via Visual Prompting and Multi-modal Chain-of-Thought (2025.acl-long)
Copied to clipboard
Yi Lu, Jiawang Cao, Yongliang Wu, Bozheng Li, Licheng Tang, Yangguang Ji, Chong Wu, Jay Wu, Wenbo Zhu
| Challenge: | Recent advances in multi-modal learning have enhanced MLLMs' ability to reason about visual content. |
| Approach: | They propose a framework that unifies multi-step multimodal reasoning with grounded visual understanding. |
| Outcome: | The proposed framework surpasses state-of-the-art methods by +6.5 gIoU and +9.2 cIou on ReasonSeg and achieves 49.7 mAP on SegInW under zero-shot settings. |
QAEval: Mixture of Evaluators for Question-Answering Task Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing QA evaluation methods struggle with open-ended and unstructured responses. |
| Approach: | They propose a hybrid framework that combines rule-based reliability with LLM-based adaptability to overcome these challenges. |
| Outcome: | The proposed framework outperforms existing models like GPT-4o and Claude-3 in accuracy and cost. |
Debiasing the Fine-Grained Classification Task in LLMs with Bias-Aware PEFT (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to mitigate label biases such as retraining, post-hoc adjustment, and parameter-efficient fine-tuning fail to address prediction propensity and discriminative ability biase. |
| Approach: | They propose a bias-aware optimization framework that incorporates two distinct label balance constraints with a PEFT strategy targeting an intermediate layer to mitigate this issue. |
| Outcome: | The proposed approach outperforms or matches the performance of full-parameter fine-tuning and LoRA, achieving superior results with lower perplexity. |
Demystifying Small Language Models for Edge Deployment (2025.acl-long)
Copied to clipboard
Zhenyan Lu, Xiang Li, Dongqi Cai, Rongjie Yi, Fangming Liu, Wei Liu, Jian Luan, Xiwen Zhang, Nicholas D. Lane, Mengwei Xu
| Challenge: | Small language models (SLMs) are a promising solution for resource-constrained devices such as smartphones and the Web of Things. |
| Approach: | They propose to use SLMs to build and optimize a set of small language models that are publicly accessible. |
| Outcome: | The proposed models outperform 7B models in general tasks, while their in-context learning capabilities remain limited and their efficiency has significant optimization potential. |
Adapt Once, Thrive with Updates: Transferable Parameter-Efficient Fine-Tuning on Evolving Base Models (2025.acl-long)
Copied to clipboard
| Challenge: | Parameter-efficient fine-tuning (PEFT) is a common method for fine- tuning large language models . however, once updated, PEFT modules suffer performance degradation on newer versions . |
| Approach: | They propose a method that enhances the PEFT module by focusing on the task-specific pattern while reducing its dependence on certain knowledge in the base model. |
| Outcome: | Experiments show that PEFT modules can maintain performance on updated models without re-tuning . the proposed approach can be used in real-world applications with large model sizes . |
Can Vision-Language Models Evaluate Handwritten Math? (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Vision-Language Models (VLMs) have significantly enhanced the ability to interpret both textual and visual data. |
| Approach: | They propose a benchmark to assess VLMs’ ability to detect, localize and correct errors in handwritten mathematical content. |
| Outcome: | The proposed benchmark covers over 2,200 handwritten math solutions from 609 manually curated problems from grades 7-12 with intentionally introduced perturbations. |
Continual Gradient Low-Rank Projection Fine-Tuning for LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Low-Rank Adaptation (LoRA) offers efficiency but constrains the model’s ability to learn new tasks and transfer knowledge due to its low-rank nature and reliance on explicit parameter constraints. |
| Approach: | They propose a training strategy that synergistically combines full and low-rank parameters and jointly updating within a unified low-ranked gradient subspace. |
| Outcome: | Extensive experiments on continual learning benchmarks show that GORP improves performance compared to state-of-the-art approaches. |
Towards Objective Fine-tuning: How LLMs’ Prior Knowledge Causes Potential Poor Calibration? (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have enabled powerful domain-specific applications through supervised fine-tuning. |
| Approach: | They propose a cognition-aware framework that applies targeted learning strategies according to the model’s prior knowledge to improve calibration. |
| Outcome: | The proposed framework significantly improves calibration while maintaining performance, achieving an average 57% reduction in ECE compared to standard fine-tuning in Llama3-8B. |
Towards Robust ESG Analysis Against Greenwashing Risks: Aspect-Action Analysis with Cross-Category Generalization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing NLP methods lack robustness against greenwashed ESG content . existing methods often extract insights that reflect misleading or exaggerated sustainability claims rather than objective ESG performance. |
| Approach: | They propose to use a dataset to improve the robustness of ESG analysis amid the prevalence of greenwashing to analyze sustainability reports. |
| Outcome: | The proposed model improves robustness against greenwashed claims rather than objective ESG performance. |
HiddenDetect: Detecting Jailbreak Attacks against Multimodal Large Language Models via Monitoring Hidden States (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on posthoc alignment techniques, but the underlying safety mechanisms within LVLMs remain unexplored. |
| Approach: | They propose a tuning-free framework that leverages internal activations to enhance safety. |
| Outcome: | The proposed framework outperforms state-of-the-art methods in detecting jailbreak attacks against large vision-language models. |
SwiLTra-Bench: The Swiss Legal Translation Benchmark (2025.acl-long)
Copied to clipboard
Joel Niklaus, Jakob Merane, Luka Nenadic, Sina Ahmadi, Yingqiang Gao, Cyrill A. H. Chevalley, Claude Humbel, Christophe Gösken, Lorenzo Tanzi, Thomas Lüthi, Stefan Palombo, Spencer Poff, Boling Yang, Nan Wu, Matthew Guillod, Robin Mamié, Daniel Brunner, Julio Pereyra, Niko Grupen
| Challenge: | In Switzerland legal translation relies on legal experts who must be both legal experts and skilled translators—creating bottlenecks and impacting effective access to justice. |
| Approach: | They propose a multilingual benchmarking system that evaluates Swiss legal translation systems based on 180K aligned Swiss legal translator pairs . they show frontier models achieve superior translation performance across all document types while specialized translation systems excel specifically in laws but under-perform in headnotes. |
| Outcome: | The proposed model outperforms specialized models in laws but underperform in headnotes. |
Two Intermediate Translations Are Better Than One: Fine-tuning LLMs for Document-level Translation Refinement (2025.acl-long)
Copied to clipboard
| Challenge: | Recent research has shown that large language models (LLMs) can enhance translation quality through self-refinement. |
| Approach: | They propose to extend translation refinement from sentence-level to document-level by using document-to-document (Doc2Doc) translations. |
| Outcome: | The proposed method improves translation quality across ten translation tasks with LLaMA-3-8B-Instruct and Mistral-Nemo-Instru. |
Circuit Compositions: Exploring Modular Structures in Transformer-Based Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in mechanistic interpretability have made progress in identifying circuits, the minimal computational subgraphs responsible for a model’s behavior on specific tasks. |
| Approach: | They propose to analyze circuits for highly compositional subtasks within a transformer-based language model to determine their modularity and how they relate to each other. |
| Outcome: | The proposed approach shows that the circuits identified exhibit notable node overlap and cross-task faithfulness. |
Can LLMs Ground when they (Don’t) Know: A Study on Direct and Loaded Political Questions (2025.acl-long)
Copied to clipboard
| Challenge: | Using large language models, interlocutors can reach mutual understanding even when they do not possess perfect knowledge. |
| Approach: | They examine whether loaded questions lead LLMs to engage in active grounding and correct false user beliefs in connection to their level of knowledge and their political bias. |
| Outcome: | The proposed model can answer direct knowledge questions and loaded questions that presuppose misinformation, while ignoring false user beliefs. |
GraphCheck: Breaking Long-Term Text Barriers with Extracted Knowledge Graph-Powered Fact-Checking (2025.acl-long)
Copied to clipboard
Yingjian Chen, Haoran Liu, Yinhong Liu, Jinxiang Xie, Rui Yang, Han Yuan, Yanran Fu, Peng Yuan Zhou, Qingyu Chen, James Caverlee, Irene Li
| Challenge: | Existing fact-checking methods that use large language models often generate subtle factual errors. |
| Approach: | They propose a fact-checking framework that uses extracted knowledge graphs to enhance text representation. |
| Outcome: | GraphCheck outperforms existing specialized fact-checkers on seven benchmarks spanning general and medical domains . Graph Neural Networks process extracted knowledge graphs as a soft prompt, enabling efficient fact- checking in a single inference call. |
SCULPT: Systematic Tuning of Long Prompts (2025.acl-long)
Copied to clipboard
Shanu Kumar, Akhila Yesantarao Venkata, Shubhanshu Khandelwal, Bishal Santra, Parag Agrawal, Manish Gupta
| Challenge: | Existing methods for prompt optimization struggle with longer, more complex ones, often risking information loss and being sensitive to small perturbations. |
| Approach: | They propose a framework that treats prompt optimization as a hierarchical tree refinement problem and uses a Critic-Actor framework to generate reflections and apply actions to refine the prompt. |
| Outcome: | The proposed framework produces more stable and interpretable prompt modifications, ensuring better generalization across tasks. |
Crab: A Novel Configurable Role-Playing LLM with Assessing Benchmark (2025.acl-long)
Copied to clipboard
Kai He, Yucheng Huang, Wenqing Wang, Delong Ran, Dongming Sheng, Junxuan Huang, Qika Lin, Jiaxing Xu, Wenqiang Liu, Mengling Feng
| Challenge: | Existing RP-LLMs employ only a single role with numerous dialogues, but Crab enables dynamic configuration of desired roles, thereby enhancing related flexibility and adaptability. |
| Approach: | They propose a Configurable Role-Playing LLM with Assessing Benchmark that combines a Role dataset curation, persona-emodying Llm construction, and comprehensive benchmark creation for RP dialogue generation. |
| Outcome: | The proposed model outperforms existing LLMs in performing fine-grained evaluations of RP while keeping dialogue per role minimal. |
Chinese SafetyQA: A Safety Short-form Factuality Benchmark for Large Language Models (2025.acl-long)
Copied to clipboard
Yingshui Tan, Boren Zheng, Baihui Zheng, Kerui Cao, Huiyun Jing, Jincheng Wei, Jiaheng Liu, Yancheng He, Wenbo Su, Xiaoyong Zhu, Bo Zheng, Kaifu Zhang
| Challenge: | Large language models have created significant safety concerns . factuality ability is crucial in determining whether they can be deployed and applied safely and compliantly within specific regions. |
| Approach: | They propose a benchmark to evaluate the factuality of large language models in China . they evaluate the models' ability to provide accurate and reliable information . |
| Outcome: | The proposed benchmark evaluates the factuality abilities of existing LLMs and compares them to LLM abilities. |
TRIDENT: Enhancing Large Language Model Safety with Tri-Dimensional Diversified Red-Teaming Data Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) excel in natural language processing tasks but are vulnerable to harmful content and being exploited for malicious purposes. |
| Approach: | They propose a framework to measure the risk coverage of alignment datasets across three dimensions: Lexical Diversity, Malicious Intent, and Jailbreak Tactics. |
| Outcome: | The proposed framework measures risk coverage across Lexical Diversity, Malicious Intent, and Jailbreak Tactics. |
Cross-Lingual Optimization for Language Transfer in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Adapting large language models to other languages often suffers from an overemphasis on English performance. |
| Approach: | They propose a cross-lingual optimization technique that efficiently transfers an English-centric LLM to a target language while preserving its English capabilities. |
| Outcome: | The proposed model outperforms SFT in acquiring target language proficiency and maintaining English performance in low-resource languages. |
CART: A Generative Cross-Modal Retrieval Framework With Coarse-To-Fine Semantic Modeling (2025.acl-long)
Copied to clipboard
Minghui Fang, Shengpeng Ji, Jialong Zuo, Hai Huang, Yan Xia, Jieming Zhu, Xize Cheng, Xiaoda Yang, Wenrui Liu, Gang Wang, Zhenhua Dong, Zhou Zhao
| Challenge: | Cross-modal retrieval tasks are used to retrieve data from one modality or another based on a query from another modality. |
| Approach: | They propose a generative cross-modal retrieval framework based on coarse-to-fine semantic modeling . they propose combining K-Means and RQ-VAE to discretize multimodal data into token sequences that support autoregressive generation. |
| Outcome: | The proposed framework achieves excellent performance and efficiency in multimodal retrieval tasks. |
MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark (2025.acl-long)
Copied to clipboard
Xiang Yue, Tianyu Zheng, Yuansheng Ni, Yubo Wang, Kai Zhang, Shengbang Tong, Yuxuan Sun, Botao Yu, Ge Zhang, Huan Sun, Yu Su, Wenhu Chen, Graham Neubig
| Challenge: | Recent advances in multimodal large language models have led to progress in tackling complex reasoning tasks that combine textual and visual information. |
| Approach: | They introduce a robust version of the Massive Multi-discipline Multimodal Understanding and Reasoning (MMMU) benchmark. |
| Outcome: | The proposed model performs lower on MMMU-Pro than on the previous benchmark, ranging from 16.8% to 26.9%. |
Cheems: A Practical Guidance for Building and Evaluating Chinese Reward Models from Scratch (2025.acl-long)
Copied to clipboard
Xueru Wen, Jie Lou, Zichao Li, Yaojie Lu, XingYu XingYu, Yuqiu Ji, Guohai Xu, Hongyu Lin, Ben He, Xianpei Han, Le Sun, Debing Zhang
| Challenge: | Existing Chinese resources are small in scale and limited to specific domains, making them insufficient for LLM post-training. |
| Approach: | They propose a Chinese-annotated reward model and a preference dataset to address this gap . they evaluate Chinese RMs on CheemsBench and construct an RM that captures human preferences . |
| Outcome: | The proposed RM achieves state-of-the-art performance on CheemsBench and CheeMePreference. |
Why Safeguarded Ships Run Aground? Aligned Large Language Models’ Safety Mechanisms Tend to Be Anchored in The Template Region (2025.acl-long)
Copied to clipboard
| Challenge: | Infilling a fixed template between the input instruction and initial model output is a common practice for existing LLMs, but it is vulnerable to inference-time jailbreak attacks. |
| Approach: | They propose to fill a fixed template between the input instruction and initial model output and to detach safety mechanisms from the template region to mitigate the risk of inference-time jailbreak attacks. |
| Outcome: | The proposed method is widespread across aligned LLMs and shows that it mitigates inference-time jailbreak vulnerabilities. |
LLaVA Steering: Visual Instruction Tuning with 500x Fewer Parameters through Modality Linear Representation-Steering (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal Large Language Models (MLLMs) enhance visual tasks by integrating visual representations into large language models. |
| Approach: | They propose a method to re-balance modalities by steering visual representations . they propose LLaVA Steering, a platform that enables rapid customization of MLLMs a component-based architecture . |
| Outcome: | The proposed model re-balances the modalities of visual representations in large language models . the model requires 500 times fewer trainable parameters than LoRA while maintaining comparable performance . |
Efficient Long Context Language Model Retrieval with Compression (2025.acl-long)
Copied to clipboard
| Challenge: | Long Context Language Models (LCLMs) are a new paradigm for Information Retrieval . however, processing large number of passages within in-context for retrieval is computationally expensive . a proposed compression approach for LCLM retrieval maximizes retrieval performance while minimizing the length of the compressed passages. |
| Approach: | They propose a new compression approach tailored to maximize retrieval performance while minimizing the length of compressed passages. |
| Outcome: | The proposed approach maximizes retrieval performance while minimizing the length of compressed passages while reducing the in-context size by 1.91. |
Ontology-Guided Reverse Thinking Makes Large Language Models Stronger on Knowledge Graph Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods rely on entity vector matching, but the purpose of the question is abstract and difficult to match with specific entities. Existing approaches rely only on entity-vector matching, and there is a problem with multi-hop reasoning. |
| Approach: | They propose a framework that constructs reasoning paths from purposes back to conditions using the KG ontology. |
| Outcome: | Experiments on the WebQSP and CWQ datasets show that ORT significantly improves the capability of large language models in knowledge graph question answering tasks (KGQA). |
Towards Omni-RAG: Comprehensive Retrieval-Augmented Generation for Large Language Models in Medical Applications (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to source planning fail to achieve this due to misalignment between the model’s expectation of the sources and their actual content. |
| Approach: | They propose a method to optimise large-scale medical knowledge models by combining multiple medical knowledge sources into one query. |
| Outcome: | The proposed method significantly improves multi-source planning performance while training a smaller model to learn source alignment. |
Predicting Turn-Taking and Backchannel in Human-Machine Conversations Using Linguistic, Acoustic, and Visual Signals (2025.acl-long)
Copied to clipboard
| Challenge: | Existing systems for human-machine conversations are limited in predicting turn-taking and backchannel actions. |
| Approach: | They propose a multi-modal face-to-face (MM-F2F) human conversation dataset . they collect and annotate over 210 hours of human conversation videos . |
| Outcome: | The proposed model achieves state-of-the-art on turn-taking and backchannel prediction tasks. |
A New Formulation of Zipf’s Meaning-Frequency Law through Contextual Diversity (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have examined Zipf's meaning-frequency law as a relationship between word frequency and the number of meanings based on contextualized word vectors . |
| Approach: | They propose to use word frequency as a relationship between word frequency and contextual diversity to examine Zipf's meaning-frequency law for a wider variety of words and corpora than previous studies have shown. |
| Outcome: | The proposed formulation gives a new interpretation of Zipf's meaning-frequency law and enables us to examine it for a wider variety of words and corpora than previous studies have shown. |
The Mirage of Model Editing: Revisiting Evaluation in the Wild (2025.acl-long)
Copied to clipboard
| Challenge: | despite near-perfect results, effectiveness of model editing in real-world applications remains unclear. |
| Approach: | They propose QAEdit and WILD to better reflect real-world use of model editing . they propose a benchmark aligned with widely used question answering datasets and a task-agnostic evaluation framework . |
| Outcome: | The proposed QAEdit benchmark and WILD evaluation framework show that current models perform worse than previously reported. |
LAQuer: Localized Attribution Queries in Content-grounded Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing attribution methods associate entire sentences with source documents, which can be overwhelming for users seeking to fact-check specific claims. |
| Approach: | They propose a task that localizes selected spans of generated output to their corresponding source spans, allowing fine-grained and user-directed attribution. |
| Outcome: | The proposed task localizes selected spans of generated output to their corresponding source spans, allowing fine-grained and user-directed attribution. |
EPO: Explicit Policy Optimization for Strategic Reasoning in LLMs via Reinforcement Learning (2025.acl-long)
Copied to clipboard
Xiaoqian Liu, Ke Wang, Yongbin Li, Yuchuan Wu, Wentao Ma, Aobo Kong, Fei Huang, Jianbin Jiao, Junge Zhang
| Challenge: | Existing methods for strategic reasoning face challenges in adaptability, scalability, and transferring strategies to new contexts. |
| Approach: | They propose an explicit policy optimization model that provides strategies in open-ended action space and can be plugged into arbitrary LLM agents to motivate goal-directed behavior. |
| Outcome: | The proposed model provides strategies in open-ended action space and can be plugged into arbitrary LLM agents to motivate goal-directed behavior. |
DCG-SQL: Enhancing In-Context Learning for Text-to-SQL with Deep Contextual Schema Link Graph (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for Text-to-SQL show little improvement compared to random selections . Existing approaches rely on intrinsic capabilities of hyper-scaled LLMs, not useful demonstrations. |
| Approach: | They propose a novel approach to effectively retrieving demonstrations and generating SQL queries by linking a question and its database schema items. |
| Outcome: | The proposed method shows consistent improvements in performance and efficiency across hyper-scaled LLMs and small LLM. |
PreP-OCR: A Complete Pipeline for Document Image Restoration and Enhanced OCR Accuracy (2025.acl-long)
Copied to clipboard
| Challenge: | Existing pipelines that combine document image restoration with semantic-aware post-OCR correction can improve text extraction from degraded images. |
| Approach: | They propose a two-stage pipeline that combines document image restoration with semantic-aware post-OCR correction to enhance both visual clarity and textual consistency. |
| Outcome: | The proposed pipeline reduces character error rates by 63.9-70.3% on 13,831 pages of real historical documents in English, French, and Spanish compared to OCR on raw images. |
Digest the Knowledge: Large Language Models empowered Message Passing for Knowledge Graph Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to augment large language models (LLMs) with external knowledge are unorganized and unorganized. |
| Approach: | They propose a method that learns a concise facts graph and encodes it into multi-level lists of texts to augment LLMs. |
| Outcome: | The proposed method improves on all 5 knowledge graph question answering datasets and offers human-level semantic explainability. |
RecLM: Recommendation Instruction Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Modern recommender systems aim to understand user-item relationships through past interactions, but their effectiveness is limited when handling sparse data or zero-shot scenarios. |
| Approach: | They propose a model-agnostic recommendation instruction-tuning paradigm that integrates large language models with collaborative filtering. |
| Outcome: | The proposed model-agnostic recommendation instruction-tuning paradigm improves performance across various settings and plug-and-play compatibility with state-of-the-art recommender systems. |
DS2-ABSA: Dual-Stream Data Synthesis with Label Refinement for Few-Shot Aspect-Based Sentiment Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for annotating data are time-consuming and labor-intensive . Existing low-resource solutions comprise data augmentation and in-context learning . |
| Approach: | They propose a dual-stream data synthesis framework for few-shot ABSA . it leverages key-point-driven and instance-driven LLMs to generate diverse data . |
| Outcome: | Extensive experiments show that DS2-ABSA outperforms existing methods . previous studies have shown that the proposed approach generates diverse data . |
MISP-Meeting: A Real-World Dataset with Multimodal Cues for Long-form Meeting Transcription and Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing systems that can recognize spoken content, extract key information, and produce concise summaries are lacking in meeting transcription and summarization. |
| Approach: | They propose a multimodal dataset that integrates information from speech, vision, and text modalities to facilitate automatic meeting transcription and summarization (AMTS). |
| Outcome: | The proposed dataset reduces the character error rate (CER) by 36.60% to 20.27% and improves speech recognition and large language models. |
Learning Together to Perform Better: Teaching Small-Scale LLMs to Collaborate via Preferential Rationale Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Prior studies have demonstrated that LLMs generate step-by-step rationales, but limited data is available to improve their performance in commercial settings due to copyright and legal issues. |
| Approach: | They propose a trainable framework that tunes a (small) LLM to generate outputs from a pool of diverse rationales that selectively improves the downstream task. |
| Outcome: | The proposed framework outperforms several trainable and prompting baselines on maths problem solving, natural language inference, and commonsense reasoning. |
MolRAG: Unlocking the Power of Large Language Models for Molecular Property Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Recent LLMs exhibit limited effectiveness on molecular property prediction task due to semantic gap between representations and natural language and lack of domain-specific knowledge. |
| Approach: | They propose a framework that integrates Chain-of-Thought reasoning for molecular property prediction. |
| Outcome: | The proposed framework outperforms pre-trained LLMs on four datasets and matches supervised methods. |
SkillAggregation: Reference-free LLM-Dependent Aggregation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods in NLP assign equal weight to all LLM judgments or are designed for specific tasks such as hallucination detection. |
| Approach: | They propose a method that learns to combine LLM judgments without additional data or ground truth to exploit the judge estimates during inference. |
| Outcome: | The proposed method outperforms Crowdlayer on all tasks and yields the best performance over all approaches on the majority of tasks. |
MasRouter: Learning to Route LLMs for Multi-Agent Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-agent systems (MAS) powered by Large Language Models (LLMs) have been demonstrated to push the boundaries of LLM capabilities, yet they often face significant costs and challenges in dynamic LLM selection. |
| Approach: | They propose a multi-agent system routing solution that integrates all components of MAS into a unified routing framework. |
| Outcome: | The proposed solution is high-performing, cost-effective, and efficient . it reduces overhead by up to 52.07 compared to current methods on HumanEval . |
Beyond Single Labels: Improving Conversational Recommendation through LLM-Powered Data Augmentation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for enhancing recommendation quality face false negatives . only one "silly cop movie" is labeled as positive, leading to suboptimal recommendations . |
| Approach: | They propose a data augmentation framework that leverages an LLM-based semantic retriever to identify diverse and semantically relevant items and filter them by a relevance scorer to remove noisy candidates. |
| Outcome: | The proposed approach improves performance on two benchmark datasets and user simulators. |
Beyond One-Size-Fits-All: Tailored Benchmarks for Efficient Evaluation (2025.acl-long)
Copied to clipboard
Peiwen Yuan, Yueqi Zhang, Shaoxiong Feng, Yiwei Li, Xinglin Wang, Jiayi Shi, Chuyi Tan, Boyuan Pan, Yao Hu, Kan Li
| Challenge: | Existing efficient methods estimate performance of models on large benchmarks, but these methods rely on the assumption that target models have high prediction consistency with source models. |
| Approach: | They propose a method that conducts customized evaluation tailored to each target model. |
| Outcome: | The proposed method reduces the MAE of estimates by 31.4% on benchmarks across 300 models. |
iQUEST: An Iterative Question-Guided Framework for Knowledge Base Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models suffer from factual inaccuracies in knowledge-intensive domains. |
| Approach: | They propose a question-guided KBQA framework that iteratively decomposes complex queries into simpler sub-questions and integrates a Graph Neural Network (GNN) to look ahead and incorporate 2-hop neighbor information at each reasoning step. |
| Outcome: | The proposed framework improves on four benchmark datasets and four LLMs. |
IRT-Router: Effective and Interpretable Multi-LLM Routing via Item Response Theory (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models have demonstrated exceptional performance across a wide range of tasks . however, selecting the optimal LLM to respond to a user query often necessitates a delicate balance between performance and cost. |
| Approach: | They propose a multi-LLM routing framework that efficiently routes user queries to the most suitable LLM. |
| Outcome: | The proposed framework outperforms baseline methods in terms of effectiveness and interpretability. |
MLAS-LoRA: Language-Aware Parameters Detection and LoRA-Based Knowledge Transfer for Multilingual Machine Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have demonstrated strong performance even with limited parallel data. |
| Approach: | They propose a multiple language-aware LoRA knowledge transfer framework that selectively adapts LLMs to MT by transferring knowledge from a large teacher to a small student model. |
| Outcome: | The proposed framework outperforms baseline models on multilingual language pairs by +1.7 BLEU on average. |
M2RC-EVAL: Massively Multilingual Repository-level Code Completion Evaluation (2025.acl-long)
Copied to clipboard
Jiaheng Liu, Ken Deng, Congnan Liu, Jian Yang, Shukai Liu, He Zhu, Peng Zhao, Linzheng Chai, Yanan Wu, JinKe JinKe, Ge Zhang, Zekun Moore Wang, Guoan Zhang, Yingshui Tan, Bangyu Xiang, Zhaoxiang Zhang, Wenbo Su, Bo Zheng
| Challenge: | Existing repository-level code completion benchmarks focus on a limited number of languages . existing benchmarks report overall average scores of different languages ignoring fine-grained abilities . |
| Approach: | They propose to use repository-level code completion benchmarks to evaluate general code intelligence abilities across languages for existing code Large Language Models. |
| Outcome: | The proposed benchmarks improve the code completion abilities of existing LLMs by using two types of annotations on the parsed syntax tree. |
Evaluating Design Decisions for Dual Encoder-based Entity Disambiguation (2025.acl-long)
Copied to clipboard
| Challenge: | Entity disambiguation (ED) is the task of linking mentions in text to corresponding entries in a knowledge base. |
| Approach: | They propose a document-level Dual Encoder model that embeds mentions and label candidates in a shared embedding space and applies a similarity metric to predict the correct label. |
| Outcome: | The proposed model improves the disambiguation of ambiguous mentions of entities in text to their respective KB entries. |
How to Compare Things Properly? A Study of Argument Relevance in Comparative Question Answering (2025.acl-long)
Copied to clipboard
Irina Nikishina, Saba Anwar, Nikolay Dolgov, Maria Manina, Daria Ignatenko, Artem Shelmanov, Chris Biemann
| Challenge: | Comparative Question Answering (CQA) is a task that involves processing information and diverse viewpoints. |
| Approach: | They construct a dataset of arguments annotated with their relevance and use it to answer comparative questions. |
| Outcome: | The proposed dataset contains arguments annotated with their relevance and enables precise traceability and faithfulness. |
FinanceReasoning: Benchmarking Financial Numerical Reasoning More Credible, Comprehensive and Challenging (2025.acl-long)
Copied to clipboard
Zichen Tang, Haihong E, Ziyan Ma, Haoyang He, Jiacheng Liu, Zhongjun Yang, Zihua Rong, Rongjin Li, Kun Ji, Qing Huang, Xinyang Hu, Yang Liu, Qianhe Zheng
| Challenge: | Compared to existing benchmarks, FinanceReasoning provides three key advancements: (1) credibility; (2) comprehensiveness; (3) numerical precision; (4) complexity; (5) complexity; and (6) complexity. |
| Approach: | They propose a benchmark to evaluate the reasoning capabilities of large reasoning models (LRMs) in financial numerical reasoning problems. |
| Outcome: | The proposed benchmark exceeds existing benchmarks in 67.8% of financial concepts and formulas and is credible, comprehensive, and challenging. |
Controllable Style Arithmetic with Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for linguistic style control lack fine-grained control, require extensive computation, or introduce significant latency. |
| Approach: | They propose a parameter-space approach that extracts style-specific representations by analyzing parameter differences between models trained on contrasting styles and incorporates them into a model with precise control over style intensity. |
| Outcome: | The proposed approach achieves three key capabilities while achieving optimal computational efficiency. |
Masks Can be Learned as an Alternative to Experts (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study shows that sparse activation techniques can reduce inference performance without sacrificing performance. |
| Approach: | They propose to sparsify a pre-trained dense large language model into a mixture-of-experts architecture for faster inference. |
| Outcome: | The proposed approach is more efficient than one-shot sparsification techniques . it achieves 97% performance retention on downstream tasks with only 50% of parameters activated . |
Program Synthesis Benchmark for Visual Programming in XLogoOnline Environment (2025.acl-long)
Copied to clipboard
| Challenge: | Large language and multimodal models have shown remarkable success on various benchmarks focused on specific skills such as general-purpose programming, math word problem-solving, and visual question answering. |
| Approach: | They propose a program synthesis benchmark based on real-world programming tasks . they propose 'fine-tuning pipeline' to boost performance of large language models . |
| Outcome: | The proposed model outperforms existing models on tasks that require a combination of skills on visual programming and programming. |
Removal of Hallucination on Hallucination: Debate-Augmented RAG (2025.acl-long)
Copied to clipboard
| Challenge: | erroneous or biased retrieval can mislead generation, compounding hallucinations. |
| Approach: | They propose a framework that integrates multi-agent debates into retrieval and generation stages to improve retrieval reliability. |
| Outcome: | The proposed framework improves retrieval reliability, reduces hallucinations and significantly improves overall factual accuracy. |
CodeDPO: Aligning Code Models with Self Generated and Verified Source Code (2025.acl-long)
Copied to clipboard
| Challenge: | Existing training methods for code generation do not improve code correctness and efficiency. |
| Approach: | They propose a framework that integrates preference learning into code generation to improve code correctness and efficiency. |
| Outcome: | The proposed framework improves code correctness and efficiency by integrating preference learning into code generation. |
ProxAnn: Use-Oriented Evaluations of Topic Models and Document Clustering (2025.acl-long)
Copied to clipboard
| Challenge: | Topic models and document clustering evaluations often use automated metrics that align poorly with human preferences or require expert labels that are intractable to scale. |
| Approach: | They propose a protocol for evaluating topic models and document clustering evaluations that uses crowdworker annotations to validate automated proxies. |
| Outcome: | The proposed protocol is scalable and easy to adapt to an LLM prompt. |
BOOKWORLD: From Novels to Interactive Agent Societies for Story Creation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have enabled social simulation through multi-agent systems. |
| Approach: | They propose a system for constructing and simulating book-based multi-agent societies that simulates established fictional worlds and characters. |
| Outcome: | The proposed system generates high-quality stories while maintaining fidelity to the source books, surpassing previous methods with a win rate of 75.36%. |
Quantifying Lexical Semantic Shift via Unbalanced Optimal Transport (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for detecting semantic change only measure the level of individual usage instances. |
| Approach: | They propose to use unbalanced optimal transport to capture semantic change through excess and deficit in the alignment between usage instances. |
| Outcome: | The proposed method captures semantic change through excess and deficit in the alignment between usage instances. |
Agentic Reward Modeling: Integrating Human Preferences with Verifiable Correctness Signals for Reliable Reward Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Existing reward models focus on human preferences, neglecting verifiable correctness signals. |
| Approach: | They propose a reward system that combines human preference rewards with verifiable correctness signals to provide reliable rewards. |
| Outcome: | The proposed reward agent significantly outperforms vanilla reward models on benchmarks and inference-time best-of-n searches on real-world tasks. |
Adaptive and Robust Translation from Natural Language to Multi-model Query Languages (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-model databases and polystore systems are increasingly studied for managing multi-model data holistically. |
| Approach: | They propose an adaptive Text-to-MMQL framework that includes a schema embedding module and an MMQl representation strategy to generate concise intermediate query formats with error correction in generated queries. |
| Outcome: | The proposed framework achieves over 9% accuracy improvement over baseline methods. |
SAKE: Steering Activations for Knowledge Editing (2025.acl-long)
Copied to clipboard
| Challenge: | Large Langue Models memorize facts, but they suffer from several limitations . Using a single input prompt is insufficient to capture the complexity of the knowledge scope affected by edits. |
| Approach: | They propose a steering activation method that models a fact to be edited as a distribution rather than a single prompt. |
| Outcome: | The proposed method can perform more robust edits than existing methods. |
Middle-Layer Representation Alignment for Cross-Lingual Transfer in Fine-Tuned LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Effective cross-lingual transfer is hindered by performance gaps and the scarcity of fine-tuning data in many languages. |
| Approach: | They propose a middle-layer alignment objective integrated into task-specific training to improve cross-lingual transfer across languages. |
| Outcome: | The proposed method improves cross-lingual transfer to lower-resource languages and can be merged with existing modules without full re-training. |
Can External Validation Tools Improve Annotation Quality for LLM-as-a-Judge? (2025.acl-long)
Copied to clipboard
| Challenge: | Pairwise feedback is widely used to evaluate and provide feedback to large language models (LLMs). |
| Approach: | They propose a tool-using agentic system to provide higher quality feedback on three challenging response domains: long-form factual, math and code tasks. |
| Outcome: | The proposed system can provide higher quality pairwise comparisons on three domains, independent of the LLM’s internal knowledge and biases. |
One for All: Update Parameterized Knowledge Across Multiple Models with Once Edit (2025.acl-long)
Copied to clipboard
Weitao Ma, Xiyuan Du, Xiaocheng Feng, Lei Huang, Yichong Huang, Huiyi Zhang, Xiaoliang Yang, Baohang Li, Xiachong Feng, Ting Liu, Bing Qin
| Challenge: | Existing methods for modifying large language models focus on individual models, resulting in errors and hallucinations. |
| Approach: | They propose an ensemble-based approach that employs a plug-in model as the editing module and a dynamic weight mechanism to enhance its effectiveness. |
| Outcome: | The proposed approach outperforms existing methods while achieving superior editing efficiency. |
VLMInferSlow: Evaluating the Efficiency Robustness of Large Vision-Language Models as a Service (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies evaluate efficiency robustness of vision-language models under unrealistic assumptions, requiring access to model architecture and parameters. |
| Approach: | They propose a novel approach to evaluate VLM efficiency robustness in a realistic black-box setting. |
| Outcome: | The proposed approach generates adversarial images with imperceptible perturbations, increasing the computational cost by up to 128.47%. |
The Alternative Annotator Test for LLM-as-a-Judge: How to Statistically Justify Replacing Human Annotators with LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are widely used in NLP research but there is no standard or rigorous procedure to determine whether they can replace humans. |
| Approach: | They propose a statistical procedure that requires only a modest subset of annotated examples to justify using LLM annotations. |
| Outcome: | The proposed procedure compares LLMs with open-source LLM annotators and judges on ten language and vision-language tasks. |
CrisisTS: Coupling Social Media Textual Data and Meteorological Time Series for Urgency Classification (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on fusion of texts and tabular-based time series to improve performance of NLP applications have focused on coupling texts with tabular time series. |
| Approach: | They propose a multimodal and multilingual dataset for urgency classification that allows for temporal and location alignment even in the absence of location mention in the text. |
| Outcome: | The proposed dataset outperforms text-only models in many applications while ensuring model generalizability. |
How to Mitigate Overfitting in Weak-to-strong Generalization? (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that weak-to-strong generalization significantly improves PGR compared to naive weak- to-strong . superalignment refers to how humans can align models on tasks beyond human ability to evaluate . |
| Approach: | They propose a framework that elicits the capabilities of strong models through weak supervisors . they propose 'superalignment' to ensure that strong models align with supervisors' intentions . |
| Outcome: | The proposed framework significantly improves quality of supervision signals and quality of input questions compared to naive weak-to-strong generalization . |
Com2 : A Causal-Guided Benchmark for Exploring Complex Commonsense Reasoning in Large Language Models (2025.acl-long)
Copied to clipboard
Kai Xiong, Xiao Ding, Yixin Cao, Yuxiong Yan, Li Du, Yufei Zhang, Jinglong Gao, Jiaqian Liu, Bing Qin, Ting Liu
| Challenge: | Existing works focus on complex tasks like math and code, while complex commonsense reasoning remains underexplored due to its uncertainty and lack of structure. |
| Approach: | They propose to build a benchmark for large language models based on complex commonsense reasoning based upon causal event graphs and causal theory. |
| Outcome: | The proposed benchmark combines a complex commonsense reasoning benchmark with a detective story to achieve a more challenging subset. |
Dynamic Head Selection for Neural Lexicalized Constituency Parsing (2025.acl-long)
Copied to clipboard
| Challenge: | Lexicalized parsing has traditionally been neglected in favor of unlexicalized, span-based methods. |
| Approach: | They propose a latent lexicalization framework that dynamically infers lexicals from data without relying on predefined head-finding rules. |
| Outcome: | The proposed model learns lexical dependencies directly from data, offering greater adaptability across languages and datasets. |
My Words Imply Your Opinion: Reader Agent-Based Propagation Enhancement for Personalized Implicit Emotion Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Current studies personalize emotion analysis by focusing on the author but neglect the impact of the intended reader on implicit emotional feedback. |
| Approach: | They propose a model which incorporates reader feedback into implicit emotion analysis (IEA) they use large language models to create reader agents to simulate reader feedback . |
| Outcome: | The proposed model outperforms state-of-the-art models in a text-centric environment. |
EvolveBench: A Comprehensive Benchmark for Assessing Temporal Awareness in LLMs on Evolving Knowledge (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have explored how LLMs perceive time, but they often overlook the critical aspect of knowledge utilization. |
| Approach: | They propose a benchmark that evaluates temporal competence along five key dimensions: Cognition, Awareness, Trustworthiness and reasoning. |
| Outcome: | EvolveBench measures temporal competence along five key dimensions: Cognition, Awareness, Trustworthiness, Understanding and reasoning. |
Enabling LLM Knowledge Analysis via Extensive Materialization (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have majorly advanced NLP and AI, and a major success factor is their internalized factual knowledge. |
| Approach: | They propose a method to comprehensively materialize an LLM’s factual knowledge through recursive querying and result consolidation. |
| Outcome: | The proposed method provides constructive insights into the scope and structure of LLM knowledge (or beliefs) it provides scale, accuracy, bias, cutoff and consistency at the same time. |
Rhythm Controllable and Efficient Zero-Shot Voice Conversion via Shortcut Flow Matching (2025.acl-long)
Copied to clipboard
Jialong Zuo, Shengpeng Ji, Minghui Fang, Mingze Li, Ziyue Jiang, Xize Cheng, Xiaoda Yang, Chen Feiyang, Xinyu Duan, Zhou Zhao
| Challenge: | Existing methods focus on disentangling speakers and content, while others focus on preserving the source's prosody. |
| Approach: | They propose a rhythm-controllable and efficient zero-shot voice conversion model that transforms the source speaker’s timbre into an unseen one while retaining speech content. |
| Outcome: | The proposed model adapts the linguistic content duration to the desired speaking style, facilitating the transfer of the target speaker’s rhythm. |
Llama See, Llama Do: A Mechanistic Perspective on Contextual Entrainment and Distraction in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | *contextual entrainment* occurs across a wide range of language models (LMs) and prompt settings. |
| Approach: | They hypothesize that there is a circuit of attention heads that corresponds to the phenomenon *contextual entrainment* . when they "turn off" these heads, the effect of contextual entraining is significantly attenuated. |
| Outcome: | The proposed method shows that LMs assign higher logits to tokens that have previously appeared in the context prompt, even for random tokens. |
CritiQ: Mining Data Quality Criteria from Human Preferences (2025.acl-long)
Copied to clipboard
Honglin Guo, Kai Lv, Qipeng Guo, Tianyi Liang, Zhiheng Xi, Demin Song, Qiuyinzhe Zhang, Yu Sun, Kai Chen, Xipeng Qiu, Tao Gui
| Challenge: | Existing methods to train language models rely on manual design, perplexity, or careful prompt engineering. |
| Approach: | They propose a method that automatically mines criteria from human preferences for data quality with only 30 human-annotated pairs and performs efficient data selection. |
| Outcome: | The proposed method improves on human-annotated test sets and shows high accuracy on code, math, and logic domains. |
Theoretical Guarantees for Minimum Bayes Risk Decoding (2025.acl-long)
Copied to clipboard
| Challenge: | Minimum Bayes Risk (MBR) decoding is a decision rule used to generate sequences from autoregressive probability models (e.g., LLMs). |
| Approach: | They propose to use minimum bayes risk (MBR) decoding to optimize output selection by maximizing expected utility value of an underlying human distribution. |
| Outcome: | The proposed method is effective even though the language space Y is larger than the hypothesis set. |
Mutual-Taught for Co-adapting Policy and Reward Models (2025.acl-long)
Copied to clipboard
Tianyuan Shi, Canbin Huang, Fanqi Wan, Longguang Zhong, Ziyi Yang, Weizhou Shen, Xiaojun Quan, Ming Yan
| Challenge: | Experimental results show that this iterative approach leads to consistent improvements in both the policy model and reward model. |
| Approach: | They propose a method that iteratively improves both the policy model and reward model without requiring additional human annotation. |
| Outcome: | The proposed method improves both the policy model and reward model without human annotation. |
Enhancing Cross-Lingual Transfer through Reversible Transliteration: A Huffman-Based Approach for Low-Resource Languages (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models demonstrate cross-lingual transfer capabilities, but these capabilities often fail to extend to low-resource languages, especially those utilizing non-Latin scripts. |
| Approach: | They propose to combine character transliteration with Huffman coding to create a complete transliterations framework that can be extended to other low-resource languages. |
| Outcome: | The proposed framework reduces storage requirements and improves accuracy and accuracy across multiple downstream tasks while maintaining performance on high-resource languages. |
Unmasking Style Sensitivity: A Causal Analysis of Bias Evaluation Instability in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to assess social biases in natural language processing models show unexpected instability when input texts undergo minor stylistic changes. |
| Approach: | They conduct a comprehensive analysis of how style transformations impact bias evaluation results . they find formal style transformation significantly affects bias scores . larger models show greater sensitivity to stylistic variations, they find . |
| Outcome: | The proposed method fails to detect appearance bias, sexual orientation bias, religious bias and religious bias in large language models. |
MockConf: A Student Interpretation Dataset: Analysis, Word- and Span-level Alignment and Baselines (2025.acl-long)
Copied to clipboard
| Challenge: | Existing parallel corpora of translated texts fail to model long-range interactions between speech segments or specific types of divergences. |
| Approach: | They propose to use MockConf to analyze simultaneous interpreting and to develop a student interpretation dataset that was collected from Mock Conferences. |
| Outcome: | The proposed dataset contains 7 hours of recordings in 5 European languages, transcribed and aligned at the level of spans and words. |
BMIKE-53: Investigating Cross-Lingual Knowledge Editing with In-Context Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Using a benchmark for cross-lingual knowledge editing, knowledge editing is underexplored. |
| Approach: | They propose a benchmark for cross-lingual in-context knowledge editing that spans 53 languages and three KE datasets. |
| Outcome: | The proposed benchmark systematically evaluates cross-lingual knowledge editing (IKE) under zero-shot, one-shot and few-shot setups. |
What Matters in Evaluating Book-Length Stories? A Systematic Study of Long Story Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | a new study examines the effectiveness of automated evaluations of book-length stories . aggregation-based and summary-based evaluations excel in detail assessment, the study finds . |
| Approach: | They propose a system for automatic evaluation of book-length stories based on human-centered criteria . they propose aggregation-based and summary-based evaluations to improve accuracy . |
| Outcome: | The proposed evaluation criteria outperforms commercial models like GPT-4o in evaluating human-written or machine-generated stories. |
PROPER: A Progressive Learning Framework for Personalized Large Language Models with Group-Level Adaptation (2025.acl-long)
Copied to clipboard
| Challenge: | Personalized large language models (LLMs) aim to tailor outputs to user preferences . however, user data is typically sparse, making it challenging to adapt LLMs to specific user patterns. |
| Approach: | They propose a progressive learning framework that groups users based on preferences and adapts LLMs in stages. |
| Outcome: | The proposed approach outperforms SOTA models across multiple tasks. |
Enhancing Event-centric News Cluster Summarization via Data Sharpening and Localization Insights (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on text summarization approaches are approaching or exceeding human excellence . |
| Approach: | They propose a framework that optimizes the balance between information volume and entropy in input texts. |
| Outcome: | The proposed framework optimizes information volume and entropy in input texts, achieving notable improvements in localized contexts. |
MMBoundary: Advancing MLLM Knowledge Boundary Awareness through Reasoning Step Confidence Calibration (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods calibrate model confidence on entire response, which leads to incorrect answers with high confidence. |
| Approach: | They propose a framework that advances the knowledge boundary awareness of multimodal large language models through reasoning step confidence calibration. |
| Outcome: | Empirical results show that the proposed framework outperforms existing methods across domains and metrics. |
LIFBench: Evaluating the Instruction Following Performance and Stability of Large Language Models in Long-Context Scenarios (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks rarely focus on instruction-following in long-context scenarios or stability on different inputs. |
| Approach: | They propose a scalable dataset to evaluate LLMs’ instruction-following capabilities and stability across long contexts. |
| Outcome: | The proposed method evaluates LLMs’ instruction-following capabilities and stability across long contexts. |
Aligning Large Language Models to Follow Instructions and Hallucinate Less via Effective Data Filtering (2025.acl-long)
Copied to clipboard
Shuzheng Si, Haozhe Zhao, Gang Chen, Cheng Gao, Yuzhuo Bai, Zhitong Wang, Kaikai An, Kangyang Luo, Chen Qian, Fanchao Qi, Baobao Chang, Maosong Sun
| Challenge: | Existing studies show that training LLMs on data containing unfamiliar knowledge during instruction tuning can encourage hallucinations. |
| Approach: | They propose a framework that measures how familiar the LLM is with instruction data and introduce an expert-aligned reward model to ensure the quality of selected samples. |
| Outcome: | The proposed framework reduces hallucinations while maintaining a competitive ability to follow instructions. |
One-Shot is Enough: Consolidating Multi-Turn Attacks into Efficient Single-Turn Prompts for LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | a novel framework for consolidating multi-turn adversarial “jailbreak” prompts into single-turn queries is presented in a journal of computational linguistics. |
| Approach: | They propose a framework for consolidating adversarial “jailbreak” prompts into single-turn queries. |
| Outcome: | The proposed framework outperforms the original multi-turn attacks by up to 17.5 % in absolute ASR . it reduces token usage by more than half on average, and provides a powerful tool for large-scale red-teaming . |
RAEmoLLM: Retrieval Augmented LLMs for Cross-Domain Misinformation Detection Using In-Context Learning Based on Emotional Information (2025.acl-long)
Copied to clipboard
| Challenge: | Current methods for cross-domain misinformation detection focus on in-domain tasks and do not incorporate significant sentiment and emotion features. |
| Approach: | They propose a retrieval augmented (RAG) LLM framework that incorporates affective information into retrieval databases. |
| Outcome: | The proposed framework improves on three misinformation benchmarks. |
Task-Specific Information Decomposition for End-to-End Dense Video Captioning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for label assignment fail to ensure relevance of localization information to descriptions. |
| Approach: | They propose a decomposed dense video captioning framework that derives localization and captioning queries from event queries, enabling task-specific representations while maintaining inter-task collaboration. |
| Outcome: | Experiments on YouCook2 and ActivityNet Captions show that the proposed framework achieves state-of-the-art performance. |
CalibraEval: Calibrating Prediction Distribution to Mitigate Selection Bias in LLMs-as-Judges (2025.acl-long)
Copied to clipboard
| Challenge: | Empirical evaluations of large language models demonstrate that they improve performance in a wide range of tasks. |
| Approach: | They propose a label-free method for mitigating selection bias during inference by reformulating debiasing as an optimization task. |
| Outcome: | The proposed method mitigates selection bias and improves performance compared to existing methods. |
Explaining Matters: Leveraging Definitions and Semantic Expansion for Sexism Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing tools for sexism detection fail to capture subtle distinctions within sexist content, limiting their practical applicability. |
| Approach: | They propose two techniques to address class imbalance and nuanced nature of sexist language . definition-based data augmentation leverages category-specific definitions to generate semantically-aligned examples . |
| Outcome: | The proposed techniques improve accuracy across all tasks and improve reliability. |
Private Memorization Editing: Turning Memorization into a Defense to Strengthen Data Privacy in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) memorize and therefore, among huge amounts of uncontrolled data, may memorize Personally Identifiable Information (PII). |
| Approach: | They propose a method that uses a model knowledge to memorize PII from training data to mitigate the memorization of PI I. |
| Outcome: | The proposed method reduces the number of leaked PIIs in a number of configurations while making it more robust against privacy Training Data Extraction attacks. |
PhysReason: A Comprehensive Benchmark towards Physics-Based Reasoning (2025.acl-long)
Copied to clipboard
Xinyu Zhang, Yuxuan Dong, Yanrui Wu, Jiaxing Huang, Chengyou Jia, Basura Fernando, Mike Zheng Shou, Lingling Zhang, Jun Liu
| Challenge: | Large language models demonstrate remarkable capabilities across various domains, including mathematics and logic reasoning. |
| Approach: | They propose a physics-based reasoning benchmark that includes physics theorems and constraints and a Physics Solution Auto Scoring Framework to evaluate physics based reasoning in large language models. |
| Outcome: | The proposed framework enables models to achieve less than 60% on answer-level evaluation, with performance dropping from knowledge questions (75.11%) to hard problems (31.99%). |
Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information (2025.acl-long)
Copied to clipboard
| Challenge: | Temporal Heads are attention heads that primarily handle temporal knowledge. |
| Approach: | They discover Temporal Heads, specific attention heads that primarily handle temporal knowledge, through circuit analysis. |
| Outcome: | The proposed models can handle temporal knowledge without compromising time-invariant and question-answering performances. |
Velocitune: A Velocity-based Dynamic Domain Reweighting Method for Continual Pre-training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to optimise pretraining performance have not addressed the complexities of domain-adaptive continual pretraining. |
| Approach: | They propose a framework that dynamically assesses learning velocity and adjusts data proportions accordingly, favouring slower learning domains while de-emphasising faster learning ones. |
| Outcome: | The proposed framework achieves performance gains in math and code reasoning tasks and command-line generation benchmarks. |
Sheep’s Skin, Wolf’s Deeds: Are LLMs Ready for Metaphorical Implicit Hate Speech? (2025.acl-long)
Copied to clipboard
| Challenge: | specialized models fail to detect implicit hate speech due to its indirectly expressed hateful intent . advanced LLMs often misinterpret metaphorical implicit hate content, resulting in its propagation . |
| Approach: | They propose a Jailbreaking strategy and Energy-based Constrained Decoding techniques to detect implicit hate speech in large language models. |
| Outcome: | The proposed model can generate metaphorical implicit hate speech, but it fails to detect it effectively. |
Neuron-Level Sequential Editing for Large Language Models (2025.acl-long)
Copied to clipboard
Houcheng Jiang, Junfeng Fang, Tianyu Zhang, Baolong Bi, An Zhang, Ruipeng Wang, Tao Liang, Xiang Wang
| Challenge: | Existing model editing methods focus on single-round editing and often face significant challenges in sequential model editing. |
| Approach: | They propose a model editing method that optimizes the target layer’s hidden states using the model’s original weights to prevent model failure. |
| Outcome: | The proposed method outperforms existing model editing methods and is available on the open-source platform 4open.science. |
Automatic Expert Discovery in LLM Upcycling via Sparse Interpolated Mixture-of-Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Sparse Interpolated Mixture-of-Experts (SIMoE) instruction-tuning is an end-to-end algorithm designed to fine-tune a dense pre-trained Large Language Model (LLM) into a MoE-style model that possesses capabilities in multiple specialized domains. |
| Approach: | They propose an algorithm to fine-tune a dense pre-trained Large Language Model into a MoE-style model that possesses capabilities in multiple specialized domains. |
| Outcome: | The proposed algorithm achieves state-of-the-art on common instruction-tuning benchmarks while maintaining an optimal performance-compute trade-off compared to baselines. |
SimulS2S-LLM: Unlocking Simultaneous Inference of Speech LLMs for Speech-to-Speech Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Simultaneous speech translation (SST) outputs translations in parallel with streaming speech input, balancing translation quality and latency. |
| Approach: | They propose to train LLMs offline and employ a test-time policy to guide simultaneous inference by extracting boundary-aware speech prompts that allow it to be better matched with text input data. |
| Outcome: | The proposed model trains speech LLMs offline and employs a test-time policy to guide simultaneous inference. |
VoxEval: Benchmarking the Knowledge Understanding Capabilities of End-to-End Spoken Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing question-answering benchmarks fail to evaluate SLMs’ knowledge understanding due to their inability to support end-to-end speech evaluation and account for varied input audio conditions. |
| Approach: | They propose a new question-answering benchmark that assesses SLMs’ knowledge understanding through pure speech interactions. |
| Outcome: | The proposed benchmark maintains speech format for both inputs and outputs, evaluates model robustness across diverse input audio conditions, and pioneers the assessment of complex tasks like mathematical reasoning in spoken format. |
RetroLLM: Empowering Large Language Models to Retrieve Fine-grained Evidence within Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods rely on separate retrievers to fetch top-k text chunks for generating evidence, and they lack joint optimization. |
| Approach: | They propose a framework that integrates retrieval and generation into a single, auto-regressive process, enabling LLMs to directly generate fine-grained evidence from the corpus with constrained decoding. |
| Outcome: | Extensive experiments on five open-domain QA datasets demonstrate the proposed framework’s superior performance across both in-domain and out-of-domain tasks. |
The Role of Deductive and Inductive Reasoning in Large Language Models (2025.acl-long)
Copied to clipboard
Chengkun Cai, Xu Zhao, Haoliang Liu, Zhongyu Jiang, Tianfang Zhang, Zongkai Wu, Jenq-Neng Hwang, Lei Li
| Challenge: | Large Language Models (LLMs) have demonstrated impressive capabilities in reasoning tasks, yet their reliability in problem-solving remains debatable. |
| Approach: | They propose a framework that integrates both deductive and inductive reasoning approaches to enhance LLM reasoning by progressively adapting its reasoning pathways based on problem complexity. |
| Outcome: | The proposed framework achieves 70.3% accuracy on AIW, compared to 62.2% for Tree of Thought, while maintaining lower computational costs. |
Disentangling the Roles of Representation and Selection in Data Pruning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for data pruning involve many different design choices, which have not been systematically studied. |
| Approach: | They decompose data pruning into two key components: data representation and selection algorithm . theoretical and empirical results highlight crucial role of representations . |
| Outcome: | The proposed method can be used to train models with less data. |
FRACTAL: Fine-Grained Scoring from Aggregate Text Labels (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies indicate that finer sentence or span-level labels provide more accurate and interpretable feedback for LLM optimization. |
| Approach: | They propose a suite of models to disaggregate response-level labels into sentence-level (pseudo-)labels through Multiple Instance Learning and Learning from Label Proportions (LLP) formulations. |
| Outcome: | The proposed model can reach 93% of the performance of the fully supervised baseline while requiring only around 10% of the gold labels. |
ACT: Knowledgeable Agents to Design and Perform Complex Tasks (2025.acl-long)
Copied to clipboard
Makoto Nakatsuji, Shuhei Tateishi, Yasuhiro Fujiwara, Ayaka Matsumoto, Narichika Nomoto, Yoshihide Sato
| Challenge: | Large language models (LLMs) have rapidly evolved and become pervasive in everyday human life. |
| Approach: | They propose a method called knowledgeable Agents to design and perform Complex Tasks where agents independently manage their knowledge and tasks while collaboratively design the complex task into a more comprehensible form. |
| Outcome: | The proposed method outperforms existing methods in solving complex tasks. |
Logical forms complement probability in understanding language model (and human) performance (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on LLMs have shown that they perform well on logical reasoning problems, but there is still a lack of fine-grained understanding of the logical forms. |
| Approach: | They propose a dataset of hypothetical and disjunctive syllogisms in propositional and modal logic and use it as the testbed for understanding LLM performance. |
| Outcome: | The proposed model performs well on proposi-tional and modal logics, but does it exhibit preferences for certain argument forms? |
Length Controlled Generation for Black-box LLMs (2025.acl-long)
Copied to clipboard
Yuxuan Gu, Wenjie Wang, Xiaocheng Feng, Weihong Zhong, Kun Zhu, Lei Huang, Ting Liu, Bing Qin, Tat-Seng Chua
| Challenge: | Existing length control methods involve fine-tuning the parameters of LLMs, which is inefficient and suboptimal for practical use. |
| Approach: | They propose an iterative sampling framework that regulates LLMs to generate length-constrained text without modifying the underlying parameters. |
| Outcome: | The proposed method achieves 100% success rates on Llama3.1 tasks with minimal additional computational overhead. |
Improving Contextual Faithfulness of Large Language Models via Retrieval Heads-Induced Optimization (2025.acl-long)
Copied to clipboard
Lei Huang, Xiaocheng Feng, Weitao Ma, Yuchun Fan, Xiachong Feng, Yangfan Ye, Weihong Zhong, Yuxuan Gu, Baoxin Wang, Dayong Wu, Guoping Hu, Bing Qin
| Challenge: | Existing frameworks for retrieval-augmented large language models (LLMs) are lacking in LFQA faithfulness testing. |
| Approach: | They propose a framework to teach retrieval-augmented large language models to explicitly discriminate between faithful and unfaithful generations. |
| Outcome: | The proposed framework outperforms GPT-4o in LFQA scenarios and outperformed existing benchmarks. |
Global Eye: Breaking the “Fixed Thinking Pattern” during the Instruction Expansion Process (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods focus on constructing multi-perspective prompts to expand instructions, overlooking the “Fixed Thinking Pattern” issue of Large Language Models. |
| Approach: | They propose a method that analyzes the statistical characteristics of newly generated instructions and updates the prompts after a fixed number of instruction expansions. |
| Outcome: | The proposed method surpasses open-source LLMs and GPT3.5 in several metrics. |
On Synthesizing Data for Context Attribution in Question Answering (2025.acl-long)
Copied to clipboard
Gorjan Radevski, Kiril Gashteovski, Shahbaz Syed, Christopher Malon, Sebastien Nicolas, Chia-Chien Hung, Timo Sztyler, Verena Heußer, Wiem Ben Rim, Masafumi Enomoto, Kunihiro Takeoka, Masafumi Oyamada, Goran Glavaš, Carolin Lawrence
| Challenge: | Large Language Models (LLMs) have a tendency to hallucinate, resulting in false or misleading answers. |
| Approach: | They propose a novel generative strategy for synthesizing context attribution data. |
| Outcome: | The proposed approach is highly effective for fine-tuning small LMs for context attribution in different QA tasks and domains. |
TST: A Schema-Based Top-Down and Dynamic-Aware Agent of Text-to-Table Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to extract text content based on static table structures neglect to deal with precise inner-document evidence extraction and dynamic information such as multiple entities and events. |
| Approach: | They propose a dynamic content extraction agent framework that uses type recognition to extract context evidences with the conduction of domain schema sequentially. |
| Outcome: | The proposed framework exhibits state-of-the-art (SOTA) performance on a large dataset. |
EventRAG: Enhancing LLM Generation with Event Knowledge Graphs (2025.acl-long)
Copied to clipboard
Zairun Yang, Yilin Wang, Zhengyan Shi, Yuan Yao, Lei Liang, Keyan Ding, Emine Yilmaz, Huajun Chen, Qiang Zhang
| Challenge: | Existing approaches to text generation often neglect event structures that shape real-world narratives. |
| Approach: | They propose a framework that integrates structured event semantics with iterative retrieval and inference to enhance text generation. |
| Outcome: | Experiments on UltraDomain and MultiHopRAG show that the proposed framework outperforms baseline RAG systems in generation effectiveness, logical consistency, and multi-hop reasoning accuracy. |
Analyzing the Rapid Generalization of SFT via the Perspective of Attention Head Activation Patterns (2025.acl-long)
Copied to clipboard
| Challenge: | Currently, LLMs learn in a data-driven schema while the instructions about complex tasks are both scarce and hard to collect or construct. |
| Approach: | They employ a gradient-based method to dissect the process that the Supervised Fine-tuning Process (SFT) adapts LLMs to downstream tasks via the perspective of attention patterns. |
| Outcome: | The proposed method dissects the process that the SFT process adapts LLMs to downstream tasks via the perspective of attention patterns. |
Can’t See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs (2025.acl-long)
Copied to clipboard
Wenxuan Wang, Xiaoyuan Liu, Kuiyi Gao, Jen-tse Huang, Youliang Yuan, Pinjia He, Shuai Wang, Zhaopeng Tu
| Challenge: | Multimodal Large Language Models (MLLMs) have expanded the capabilities of traditional language models by enabling interaction through both text and images. |
| Approach: | They propose a multimodal safety awareness benchmark to evaluate MLLMs across 29 safety scenarios with 1,500 carefully curated image-prompt pairs. |
| Outcome: | The proposed model is able to identify unsafe content and avoid over-sensitivity that can hinder helpfulness. |
Mis-prompt: Benchmarking Large Language Models for Proactive Error Handling (2025.acl-long)
Copied to clipboard
| Challenge: | Current error-handling works are performed in a passive manner, with explicit error- handling instructions. |
| Approach: | They propose a new benchmark to analyze LLMs' performance on a mis-prompt benchmark and a dataset to promote further research. |
| Outcome: | The proposed benchmark shows that current LLMs show poor performance on proactive error handling, and that SFT improves on error handling instances. |
TripCraft: A Benchmark for Spatio-Temporally Fine Grained Travel Planning (2025.acl-long)
Copied to clipboard
Soumyabrata Chaudhuri, Pranav Purkar, Ritwik Raghav, Shubhojit Mallick, Manish Gupta, Abhik Jana, Shreya Ghosh
| Challenge: | Existing benchmarks such as TravelPlanner and TravelPlann+ rely on semi-synthetic data and lack key real-world constraints. |
| Approach: | They propose a spatio-temporally coherent travel planning dataset incorporating real-world constraints, including public transit schedules, public events, varied attraction categories, and user personas for enhanced personalization. |
| Outcome: | The proposed dataset significantly improves meal scheduling, improving performance from 61% to 80% in the 7-day scenario. |
DualGuard: A Parameter Space Transformation Approach for Bidirectional Defense in Split-Based LLM Fine-Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing defense methods for large language model fine-tuning (LLM-FT) sacrifice task-specific performance under privacy constraints. |
| Approach: | They propose a bidirectional defense mechanism that uses a local warm-up parameter transformation to alter client-side model parameters before training. |
| Outcome: | The proposed defense mechanism outperforms current defense methods while maintaining task performance. |
Movie101v2: Improved Movie Narration Benchmark (2025.acl-long)
Copied to clipboard
| Challenge: | Automatic movie narration aims to generate video-aligned plot descriptions to assist visually impaired audiences. |
| Approach: | They propose to break down the ultimate goal of automatic movie narration into three stages . they propose a large-scale, bilingual dataset with enhanced data quality . |
| Outcome: | The proposed dataset breaks down the goal of automatic movie narration into three stages . achieving applicable movie narration is a fascinating goal that requires significant research . |
Can LLMs Evaluate Complex Attribution in QA? Automatic Benchmarking using Knowledge Graphs (2025.acl-long)
Copied to clipboard
Nan Hu, Jiaoyan Chen, Yike Wu, Guilin Qi, Hongru Wang, Sheng Bi, Yongrui Chen, Tongtong Wu, Jeff Z. Pan
| Challenge: | Attributed Question Answering (AQA) has attracted wide attention, but there are several limitations in evaluating the attributions. |
| Approach: | They propose a large-scale benchmark containing comprehensive attribution categories . they compare 25 automatic evaluators with human evaluers and tested LLM evalators . |
| Outcome: | The proposed method can compare attributions with subtle differences and provide feedback to improve them. |
Value Portrait: Assessing Language Models’ Values through Psychometrically and Ecologically Valid Items (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks rely on human annotations that are vulnerable to value-related biases. |
| Approach: | They propose a value portrait benchmark that uses items that capture real-life user-LLM interactions and a rated item based on its similarity to their own thoughts to determine reliability. |
| Outcome: | The proposed framework improves the relevance of assessment results to real-world LLM usage by allowing human subjects to rate items with similarity to their own thoughts and derived correlations between these ratings and the subjects’ actual value scores. |
FEA-Bench: A Benchmark for Evaluating Repository-Level Code Generation for Feature Implementation (2025.acl-long)
Copied to clipboard
Wei Li, Xin Zhang, Zhongxin Guo, Shaoguang Mao, Wen Luo, Guangyue Peng, Yangyu Huang, Houfeng Wang, Scarlett Li
| Challenge: | Existing benchmarks focus on standalone programming problems, such as HumanEval, MBPP, and LiveCodeBench. |
| Approach: | They propose to use large language models to evaluate their ability to perform incremental development within code repositories by collecting pull requests from 83 GitHub repositorias and using rule-based and intent-based filtering to construct task instances focused on new feature development. |
| Outcome: | The proposed benchmarks show that large language models perform significantly worse in the FEA-Bench, highlighting considerable challenges in repository-level incremental code development. |
Do not Abstain! Identify and Solve the Uncertainty (2025.acl-long)
Copied to clipboard
| Challenge: | Existing solutions rely on evasive responses when confronting uncertain scenarios. |
| Approach: | They propose a benchmark to assess LLMs' ability to recognize and address uncertainty . they generate context-aware inquiries that highlight the confusing aspect of the original query . |
| Outcome: | Experiments with ConfuseBench show that LLMs struggle to identify root cause of uncertainty and solve it. |
Decoding by Contrasting Knowledge: Enhancing Large Language Model Confidence on Edited Facts (2025.acl-long)
Copied to clipboard
| Challenge: | In-context knowledge editing (ICE) is currently the most effective method for knowledge editing, but it is constrained by the black-box modeling of LLMs and lacks interpretability. |
| Approach: | They propose a method to decode new knowledge by comparing logits with unedited knowledge to improve the accuracy of LLMs. |
| Outcome: | The proposed method improves the performance of LLaMA3-8B-instruct on MQuAKE by up to 219%. |
ImpliHateVid: A Benchmark Dataset and Two-stage Contrastive Learning Framework for Implicit Hate Speech Detection in Videos (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on text-based hate speech detection focus on video-based approaches . however, hateful content remains a persistent challenge due to the vast amount of data generated every day. |
| Approach: | They propose a novel two-stage contrastive learning framework for hate speech detection in videos . they train modality-specific encoders for audio, text, and image using contrastive loss . |
| Outcome: | The proposed framework is based on two datasets, ImpliHateVid and HateMM datasets. |
Improving Chain-of-Thought Reasoning via Quasi-Symbolic Abstractions (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work suggests using logical formalisms coupled with external symbolic solvers to solve complex tasks. |
| Approach: | They propose a framework to disentangle content from logical reasoning without a complete formalisation. |
| Outcome: | The proposed methods improve CoT-based methods by up to 8% accuracy on challenging adversarial variations on both natural language and symbolic reasoning tasks. |
Information Extraction from Visually Rich Documents using LLM-based Organization of Documents into Independent Textual Segments (2025.acl-long)
Copied to clipboard
| Challenge: | Specialized non-LLM NLP-based solutions lack reasoning and are not able to infer values not explicitly present in documents. |
| Approach: | They propose a novel LLM-based approach that organizes VRDs into localized semantic textual segments called semantic blocks. |
| Outcome: | The proposed approach outperforms the state-of-the-art on public VRD benchmarks by 1-3% in F1 scores and is resilient to document formats previously not encountered. |
Enhancing Open-Domain Task-Solving Capability of LLMs via Autonomous Tool Integration from GitHub (2025.acl-long)
Copied to clipboard
Bohan Lyu, Xin Cong, Heyang Yu, Pan Yang, Cheng Qian, Zihe Wang, Yujia Qin, Yining Ye, Yaxi Lu, Chen Qian, Zhong Zhang, Yukun Yan, Yankai Lin, Zhiyuan Liu, Maosong Sun
| Challenge: | Existing approaches lack flexibility to address diverse and ever-evolving user queries in open domains. |
| Approach: | They propose to evaluate LLMs on open-domain knowledge that requires tools to solve diverse and ever-evolving user queries. |
| Outcome: | The proposed system outperforms baselines in the open domain task-solving benchmark. |
LLMs Can Simulate Standardized Patients via Agent Coevolution (2025.acl-long)
Copied to clipboard
Zhuoyun Du, LujieZheng LujieZheng, Renjun Hu, Yuyang Xu, Xiawei Li, Ying Sun, Wei Chen, Jian Wu, Haolei Cai, Haochao Ying
| Challenge: | Training medical personnel using standardized patients (SPs) remains a complex challenge, necessitating extensive domain expertise and role-specific practice. |
| Approach: | They propose a simulated patient framework that allows patient agents to simulate diagnostic process through multi-turn dialogues. |
| Outcome: | The proposed framework improves over existing reasoning methods by more than 10% in requirement alignment and better human preference after evolving over 200 cases for 10 hours with excellent generalizability. |
Donate or Create? Comparing Data Collection Strategies for Emotion-labeled Multimodal Social Media Posts (2025.acl-long)
Copied to clipboard
| Challenge: | Accurate modeling of subjective phenomena requires data annotated with authors’ intentions. |
| Approach: | They collect study-created and genuine social media posts labeled for emotion and compare them on several dimensions, including model performance. |
| Outcome: | The results show that study-created posts are longer, rely more on text and less on images for emotion expression, and focus more on emotion-prototypical events. |
Which Demographics do LLMs Default to During Annotation? (2025.acl-long)
Copied to clipboard
Johannes Schäfer, Aidan Combs, Christopher Bagdon, Jiahui Li, Nadine Probol, Lynn Greschner, Sean Papay, Yarik Menchaca Resendiz, Aswathy Velutharambath, Amelie Wuehrl, Sabine Weber, Roman Klinger
| Challenge: | Demographics and cultural background of annotators influence the labels they assign in text annotation. |
| Approach: | They examine the attributes of human annotators LLMs inherently mimic and compare them to demographic-conditioned prompts and placebo-conditioned ones. |
| Outcome: | The proposed model incorporates demographics and cultural background into the output of the large language models (LLMs) to evaluate which attributes of human annotators LLMs inherently mimic. |
Can You Really Trust Code Copilot? Evaluating Large Language Models from a Code Security Perspective (2025.acl-long)
Copied to clipboard
| Challenge: | Existing code security benchmarks focus on one task and paradigm, such as code completion and generation, without comprehensive assessment across dimensions like secure code generation, vulnerability repair and discrimination. |
| Approach: | They propose a multi-task benchmark for comprehensive evaluation of LLM code security . they also propose VC-Judge, an improved judgment model that aligns closely with human experts . |
| Outcome: | The proposed model can evaluate LLM-generated programs for vulnerabilities in a more efficient and reliable way. |
From Sub-Ability Diagnosis to Human-Aligned Generation: Bridging the Gap for Text Length Control via MarkerGen (2025.acl-long)
Copied to clipboard
Peiwen Yuan, Chuyi Tan, Shaoxiong Feng, Yiwei Li, Xinglin Wang, Yueqi Zhang, Jiayi Shi, Boyuan Pan, Yao Hu, Kan Li
| Challenge: | Existing methods to control text length are lacking in LCTG, posing a major limitation for practical applications. |
| Approach: | They propose a plug-and-play approach that decomposes LCTG sub-abilities with human patterns as reference and performs detailed error analysis. |
| Outcome: | The proposed method significantly improves LCTG across various settings, exhibiting outstanding effectiveness and generalizability. |
AGD: Adversarial Game Defense Against Jailbreak Attacks in Large Language Models (2025.acl-long)
Copied to clipboard
Shilong Pan, Zhiliang Tian, Zhen Huang, Wanlong Yu, Zhihua Wen, Xinwang Liu, Kai Lu, Minlie Huang, Dongsheng Li
| Challenge: | Existing defenses, including post-training alignment and prompt engineering, struggle with adaptability to out-of-distribution (OOD) attacks. |
| Approach: | They propose an adversarial game-based defense method that dynamically adjusts LLMs’ internal representations to achieve a balanced trade-off between helpfulness and harmlessness. |
| Outcome: | The proposed method improves LLMs’ safety over all baselines. |
SCOP: Evaluating the Comprehension Process of Large Language Models from a Cognitive View (2025.acl-long)
Copied to clipboard
| Challenge: | despite the potential of large language models, it is difficult to fully count on them in real-world scenarios. |
| Approach: | They propose to examine how LLMs perform during the comprehension process from a cognitive perspective. |
| Outcome: | The proposed model analyzes how LLMs perform during the comprehension process from a cognitive perspective. |
Table-Critic: A Multi-Agent Framework for Collaborative Criticism and Refinement in Table Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to decompose large language models (LLMs) lack effective mechanisms to identify and correct errors in intermediate reasoning steps, leading to cascading error propagation. |
| Approach: | They propose a multi-agent framework that facilitates collaborative criticism and iterative refinement of the reasoning process until convergence to correct solutions. |
| Outcome: | The proposed framework achieves superior accuracy and error correction rates while maintaining computational efficiency and lower solution degradation rate. |
An Expanded Massive Multilingual Dataset for High-Performance Language Technologies (HPLT) (2025.acl-long)
Copied to clipboard
Laurie Burchell, Ona De Gibert Bonet, Nikolay Arefyev, Mikko Aulamo, Marta Bañón, Pinzhen Chen, Mariia Fedorova, Liane Guillou, Barry Haddow, Jan Hajič, Jindřich Helcl, Erik Henriksson, Mateusz Klimaszewski, Ville Komulainen, Andrey Kutuzov, Joona Kytöniemi, Veronika Laippala, Petter Mæhlum, Bhavitvya Malik, Farrokh Mehryary, Vladislav Mikhailov, Nikita Moghe, Amanda Myntti, Dayyán O’Brien, Stephan Oepen, Proyag Pal, Jousia Piha, Sampo Pyysalo, Gema Ramírez-Sánchez, David Samuel, Pavel Stepachev, Jörg Tiedemann, Dušan Variš, Tereza Vojtěchová, Jaume Zaragoza-Bernabeu
| Challenge: | a large number of textual data is needed to train state-of-the-art large language models. |
| Approach: | They propose a collection of monolingual and parallel corpora from the Internet Archive . they document the entire data pipeline and release the code to reproduce it . |
| Outcome: | The proposed collection of monolingual and parallel corpora is based on the HPLT v2 dataset . it includes 8T tokens covering 193 languages and 380M sentence pairs covering 51 languages . |
Scaling Text-Rich Image Understanding via Code-Guided Synthetic Multimodal Data Generation (2025.acl-long)
Copied to clipboard
Yue Yang, Ajay Patel, Matt Deitke, Tanmay Gupta, Luca Weihs, Andrew Head, Mark Yatskar, Chris Callison-Burch, Ranjay Krishna, Aniruddha Kembhavi, Christopher Clark
| Challenge: | Vision-language models struggle to understand text-rich images due to the scarcity of diverse text-only large language data. |
| Approach: | They propose a framework that leverages the coding capabilities of text-only large language models to create synthetic text-rich multimodal data. |
| Outcome: | The proposed framework can generate high-quality instruction-tuning data using Python, HTML, LaTeX and other languages. |
Hierarchical Attention Generates Better Proofs (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have shown promise in formal theorem proving, but their token-level processing often fails to capture the inherent hierarchical nature of mathematical proofs. |
| Approach: | They propose a regularization method that aligns LLMs’ attention mechanisms with mathematical reasoning structures and establishes a five-level hierarchy from foundational elements to high-level concepts. |
| Outcome: | The proposed method improves proof success rates by 2.05% on miniF2F and 1.69% on ProofNet while reducing proof complexity by 23.81% and 16.50% respectively. |
Agent-RewardBench: Towards a Unified Benchmark for Reward Modeling across Perception, Planning, and Safety in Real-World Multimodal Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal Large Language Models (MLLMs) are developing but lack external feedback . there is no clear on how to select reward models for agents . |
| Approach: | They propose a benchmark to evaluate agent reward modeling ability in MLLMs . they use multiple dimensions and real-world agent scenarios evaluation . |
| Outcome: | The proposed benchmark evaluates agent performance in multimodal large language models . it covers perception, planning, and safety with 7 scenarios and is highly difficult and high-quality . |
It’s Not Bragging If You Can Back It Up: Can LLMs Understand Braggings? (2025.acl-long)
Copied to clipboard
| Challenge: | Bragging is a pervasive social-linguistic phenomenon that reflects complex human interaction patterns. |
| Approach: | They propose to use bragging recognition, bragging explanation, and bragging generation tasks to examine bragging in large language models (LLMs) . |
| Outcome: | The proposed models can identify bragging intent, social appropriateness, and account for context sensitivity and provide new insights into how LLMs process bragging. |
A Troublemaker with Contagious Jailbreak Makes Chaos in Honest Towns (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research focuses on single-agent attacks and shared memory attacks, but real-world scenarios often involve independent memory. |
| Approach: | They propose a large-scale, multi-agent, multitopology attack evaluation framework that exploits the memory of an agent to make it more vulnerable to jailbreak attacks. |
| Outcome: | The proposed framework improves on the troublemaker makes chaos in Honest Town task with 23.51%, 18.95%, and 52.93% improvements in line, star topologies, and 100-agent settings. |
Meta-Learning Neural Mechanisms rather than Bayesian Priors (2025.acl-long)
Copied to clipboard
| Challenge: | a meta-learning approach has been proposed to integrate human-like learning biases into neural networks . a recent study suggests that meta-training on a single formal language can improve a model . |
| Approach: | They propose to integrate human-like learning biases into neural-network architectures . they use symbolic models to capture aspects of humans' basic generalisations from small data . |
| Outcome: | The proposed model can learn from a single language as much as 5000 different languages . the model can be scaled to a larger model and training datasets . |
Shifting from Ranking to Set Selection for Retrieval Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches primarily rerank top-k passages based on individual relevance, often failing to meet the information needs of complex queries in multi-hop question answering. |
| Approach: | They propose a set-wise passage selection approach and introduce SetR which explicitly identifies the information requirements of a query through Chain-of-Thought reasoning. |
| Outcome: | The proposed approach outperforms both proprietary LLM-based rerankers and open-source baselines in terms of answer correctness and retrieval quality. |
Understanding Large Language Model Vulnerabilities to Social Bias Attacks (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated remarkable linguistic capabilities across tasks . however, there is a growing concern about their potential to perpetuate social biases . |
| Approach: | They evaluate LLMs across gender, racial, and religious bias types . they also explore cross-bias and multiple-biases attacks . |
| Outcome: | The proposed models are more susceptible to gender bias attacks than racial or religious biases. |
ChatSOP: An SOP-Guided MCTS Planning Framework for Controllable LLM Dialogue Agents (2025.acl-long)
Copied to clipboard
Zhigen Li, Jianxiang Peng, Yanmeng Wang, Yong Cao, Tianhao Shen, Minghui Zhang, Linxi Su, Shang Wu, Yihang Wu, YuQian Wang, Ye Wang, Wei Hu, Jianfeng Li, Shaojun Wang, Jing Xiao, Deyi Xiong
| Challenge: | Existing models that use Large Language Models (LLMs) show superior performance in various tasks, but lack of controllability leads to unfocused conversations or task failure. |
| Approach: | They propose a standard operating procedure (SOP) framework to regulate dialogue flow by integrating Chain of Thought reasoning and supervised fine-tuning for SOP prediction. |
| Outcome: | The proposed method achieves a 27.95% improvement in action accuracy compared to baseline models based on GPT-3.5 and also shows notable gains for open-source models. |
Pixel-Level Reasoning Segmentation via Multi-turn Conversations (2025.acl-long)
Copied to clipboard
| Challenge: | Existing visual perception systems focus on region-level segmentation in single-turn dialogues . existing systems cannot reason at the pixel level and comprehend dynamic user intent . |
| Approach: | They propose a task that tracks evolving user intent via multi-turn interactions for fine-grained segmentation. |
| Outcome: | The proposed method outperforms existing baselines in segmentation and reasoning metrics. |
Fixing Distribution Shifts of LLM Self-Critique via On-Policy Self-Play Training (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models show impressive performance in a wide range of linguistic tasks, but their performance on complex reasoning tasks is still signif-icantly lower than the human level. |
| Approach: | They propose a reinforcement learning framework to synchronize the reasoning and critique capabilities of language models by using Monte Carlo sampling to give appropriate rewards to the model's critique content. |
| Outcome: | The proposed framework improves the model's reasoning and critique capabilities by 5.40 and 3.66 points, respectively, compared to the best baseline approach. |
Inferring Functionality of Attention Heads from their Parameters (2025.acl-long)
Copied to clipboard
| Challenge: | Prior work on attention heads focused on analyzing their behavior during inference for specific circuits or tasks. |
| Approach: | They propose a framework that maps the functions of attention heads without any model training or inference. |
| Outcome: | The proposed framework infers the functionality of attention heads without any model training or inference. |
Faithful and Robust LLM-Driven Theorem Proving for NLI Explanations (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work has shown that the interaction of large language models (LLMs) with theorem provers (TPs) can help verify and improve the validity of NLI explanations. |
| Approach: | They propose to use logical expressions to guide LLMs in generating structured proof sketches and to use them to improve their accuracy. |
| Outcome: | The proposed strategies improve autoformalisation, syntactic errors and explanation refinement over the state-of-the-art model. |
Revealing the Deceptiveness of Knowledge Editing: A Mechanistic Analysis of Superficial Editing (2025.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge editing algorithms are prone to generating original knowledge . despite the fact that many models achieve near-perfect performance, superficial editing remains a challenge . |
| Approach: | They propose to use "**superficial editing**" to describe the phenomenon . they investigate the internal mechanisms of the attention module and their corresponding left singular vectors . |
| Outcome: | The proposed method can modify specific knowledge in a pretrained large language model while ensuring that unrelated knowledge remains unaffected. |
Masking in Multi-hop QA: An Analysis of How Language Models Perform with Context Permutation (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-hop Question Answering (MHQA) adds layers of complexity to question answering tasks. |
| Approach: | They explore how LMs respond to multi-hop questions by permuting search results under various configurations. |
| Outcome: | The proposed model outperforms decoder-only models in MHQA tasks despite being significantly smaller in size . |
From Human Reading to NLM Understanding: Evaluating the Role of Eye-Tracking Data in Encoder-Based Models (2025.acl-long)
Copied to clipboard
| Challenge: | integrating eye-tracking features into Neural Language Models does not degrade downstream task performance, enhances alignment between model attention and human attention patterns, and compresses the embedding space. |
| Approach: | They used eye-gaze data from the Ghent Eye-Tracking Corpus to investigate how integrating knowledge of human reading behavior impacts Neural Language Models. |
| Outcome: | The proposed approach does not degrade downstream task performance, enhances alignment between model attention and human attention patterns, and compresses the embedding space. |
Optimizing Question Semantic Space for Dynamic Retrieval-Augmented Multi-hop Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing retrieval-augmented generation methods are insufficient for multi-hop question answering . however, they tend to generate hallucinations due to semantic mismatching . |
| Approach: | They propose to optimize question semantic space for dynamic retrieval-augmented multi-hop question answering by optimizing the semantic embeddings. |
| Outcome: | The proposed method outperforms existing RAG methods in both in- and out-of-domain settings. |
Insight Over Sight: Exploring the Vision-Knowledge Conflicts in Multimodal LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to mitigating vision-knowledge conflict in Large Language Models (MLLMs) are not effective and can be further scaled. |
| Approach: | They propose a framework to generate inputs to simulate and evaluate vision-knowledge conflict in Multimodal Large Language Models (MLLMs) using original images and 1,122 high-quality question-answer pairs, they propose 'a diagnostic benchmark' |
| Outcome: | The proposed framework, benchmark, and analysis contribute to the understanding and mitigation of vision-knowledge conflicts in Multimodal Large Language Models (MLLMs). |
SceneGenAgent: Precise Industrial Scene Generation with Coding Agent (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work on scene generation focuses on generating 3D scenes from textual descriptions . however, the task of generating industrial scenes with LLMs is complex and requires precise measurements and positioning . |
| Approach: | They propose an LLM-based agent for generating industrial scenes through C# code. |
| Outcome: | Experiments show that LLMs powered by SceneGenAgent exceed their original performance . the agent achieves 81.0% success rate in real-world industrial scene generation tasks . |
ToolCoder: A Systematic Code-Empowered Tool Learning Framework for Large Language Models (2025.acl-long)
Copied to clipboard
Hanxing Ding, Shuchang Tao, Liang Pang, Zihao Wei, Jinyang Gao, Bolin Ding, Huawei Shen, Xueqi Cheng
| Challenge: | Existing approaches to tool learning rely on hand-crafted prompts and natural language reasoning, making multi-step planning difficult and lacking precise error diagnosis and reflection mechanisms. |
| Approach: | They propose a framework that reformulates tool learning as a code generation task. |
| Outcome: | The proposed framework achieves superior performance in task completion accuracy and execution reliability compared to existing approaches. |
Enhancing Text Editing for Grammatical Error Correction: Arabic as a Case Study (2025.acl-long)
Copied to clipboard
| Challenge: | Text editing is a wellstudied problem for grammatical error correction (GEC) but it is not the most efficient for morphologically rich languages like Arabic. |
| Approach: | They propose a text editing approach that derives edit tags directly from data, eliminating the need for language-specific edits. |
| Outcome: | The proposed approach achieves SOTA results on Arabic and performs on par with SOTA on two other languages. |
From Isolates to Families: Using Neural Networks for Automated Language Affiliation (2025.acl-long)
Copied to clipboard
| Challenge: | linguistic affiliation of languages to a common language family is traditionally carried out manually . large-scale standardized collections of multilingual wordlists and grammatical language structures could improve this . |
| Approach: | They propose to use lexical and grammatical data to classify languages into families using neural network models. |
| Outcome: | The proposed models outperform models trained on lexical and grammatical data while combining both types of data yields even better performance. |
ELBA-Bench: An Efficient Learning Backdoor Attacks Benchmark for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing backdoor models are limited in coverage of attack, system integrity and backdoor alignment . ELBA-Bench provides over 1300 experiments encompassing 12 attack methods, 18 datasets, and 12 LLMs. |
| Approach: | They propose a framework that allows attackers to inject backdoor through parameter efficient fine-tuning or without fine-uning techniques. |
| Outcome: | ELBA-Bench provides over 1300 experiments encompassing 12 attack methods, 18 datasets, and 12 LLMs. |
Less, but Better: Efficient Multilingual Expansion for LLMs via Layer-wise Mixture-of-Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Existing large language models (LLMs) have remarkable ability in high-resource languages, but their performance in multilingual scenarios is still limited. |
| Approach: | They propose a layer-wise expert allocation algorithm to determine the appropriate number of new experts for each layer. |
| Outcome: | The proposed method outperforms the previous state-of-the-art baseline with 60% fewer experts in the single-expansion setting and 33.3% fewer in the lifelong-expanding setting. |
When Harry Meets Superman: The Role of The Interlocutor in Persona-Based Dialogue Generation (2025.acl-long)
Copied to clipboard
| Challenge: | In recent years, large language models (LLMs) have proven effective in generating coherent and contextually appropriate responses. |
| Approach: | They examine the ability of a model to adapt to the interlocutor's profile by masking or disclosing information about interlucutor . |
| Outcome: | The proposed model generalises well across topics, but struggles with unfamiliar interlocutors. |
ICR Probe: Tracking Hidden State Dynamics for Reliable Hallucination Detection in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for hallucination detection rely on static and isolated representations, overlooking their dynamic evolution across layers. |
| Approach: | They propose a method which captures the cross-layer evolution of hidden states and propose 'ICR Probe' which capture the evolution of the hidden states. |
| Outcome: | The proposed method achieves superior performance with significantly fewer parameters and ablation studies offer deeper insights into the underlying mechanism of the method, improving its interpretability. |
Revisit Self-Debugging with Self-Generated Tests for Code Generation (2025.acl-long)
Copied to clipboard
Xiancai Chen, Zhengwei Tao, Kechi Zhang, Changzhi Zhou, Xinyu Zhang, Wanli Gu, Yuanpeng He, Mengdi Zhang, Xunliang Cai, Haiyan Zhao, Zhi Jin
| Challenge: | Large language models (LLMs) have made significant advances in code generation, but they still face challenges when tackling complex programming tasks beyond their basic capabilities. |
| Approach: | They propose to integrate self-generated tests into the code generation process . they propose to use post-execution and in-exection self-debugging to mitigate test bias . |
| Outcome: | The proposed method improves the performance of large language models in code generation tasks by leveraging execution feedback from tests. |
InSerter: Speech Instruction Following with Unsupervised Interleaved Pre-training (2025.acl-long)
Copied to clipboard
Dingdong Wang, Jin Xu, Ruihang Chu, Zhifang Guo, Xiong Wang, Jincenzi Wu, Dongchao Yang, Shengpeng Ji, Junyang Lin
| Challenge: | Recent advances in speech large language models exhibit suboptimal performance in adhering to speech instructions. |
| Approach: | They propose a method to pre-train large-scale unsupervised speech-text sequences . they use text-to-speech conversion to generate textual continuations corresponding to provided speech segments . |
| Outcome: | The proposed model achieves superior or competitive results across diverse speech processing tasks. |
Exploring LLMs’ Ability to Spontaneously and Conditionally Modify Moral Expressions through Text Manipulation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on moral-related tasks based on large language models have not been conducted. |
| Approach: | They analyze behavior of Large Language Models (LLMs) among open and uncensored models and use human-annotated datasets to analyze moral-related data. |
| Outcome: | The results show that large language models can alter moral dimensions through text manipulation tasks and moral-related conditioning prompts. |
Mixture of Ordered Scoring Experts for Cross-prompt Essay Trait Scoring (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to automate essay scoring overlook critical information, authors say . evaluators often limit their performance to unseen topics, resulting in incomplete assessment perspectives. |
| Approach: | They propose a framework that integrates information from prompts and essays into an AES framework. |
| Outcome: | The proposed framework achieves state-of-the-art in cross-prompt scoring and multi-trait scoring on the ASAP++ dataset. |
Sparse Logit Sampling: Accelerating Knowledge Distillation in LLMs (2025.acl-long)
Copied to clipboard
Anshumann Anshumann, Mohd Abbas Zaidi, Akhil Kedia, Jinwoo Ahn, Taehwak Kwon, Kangwook Lee, Haejun Lee, Joohyung Lee
| Challenge: | Knowledge distillation is a cost-effective technique to distill knowledge in Large Language Models, if the teacher output logits can be pre-computed and cached. |
| Approach: | They propose an importance-sampling-based method which provides unbiased estimates, preserves the gradient in expectation, and requires storing significantly sparser logits. |
| Outcome: | The proposed method enables faster training of student models with marginal overhead (10%) compared to cross-entropy based training, while maintaining competitive performance compared with full distillation. |
Enhancing Spoken Discourse Modeling in Language Models Using Gestural Cues (2025.acl-long)
Copied to clipboard
| Challenge: | linguistic research shows that non-verbal cues, such as gestures, play a crucial role in spoken discourse. |
| Approach: | They propose to integrate gestures into language models by embedding human motion sequences into discrete gesture tokens and aligning them with text embeddables. |
| Outcome: | The proposed model improves on spoken discourse, the authors show . the study aims to improve the accuracy of discourse markers and quantifiers . |
ExploraCoder: Advancing Code Generation for Multiple Unseen APIs via Planning and Chained Exploration (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models face intrinsic limitations in coding with unseen APIs in training corpora. |
| Approach: | They propose a training-free framework that empowers LLMs to invoke multiple unseen APIs in code solution by planning a complex problem into several API invocation subtasks and experimenting with correct API usage at intermediate steps. |
| Outcome: | The proposed framework significantly improves performance for models lacking prior API knowledge, achieving 11.99% over retrieval-based approaches and 17.28% over pretraining-based methods in pass@10. |
Segment First or Comprehend First? Explore the Limit of Unsupervised Word Segmentation with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to measure word segmentation only assess the language model's understanding of the overall meaning of sentences, lacking an evaluation of the language models' understanding capabilities at a fine-grained level. |
| Approach: | They propose a framework to explore the limit of unsupervised word segmentation with Large Language Models (LLMs) they employ current mainstream LLMs to perform word segmentations across multiple languages . |
| Outcome: | The proposed method improves on existing methods and combines the advanced pattern recognition capabilities of Aho-Corasick automata with the deep insights of well-pretrained LLMs. |
RUBY: An Effective Framework for Multi-Constraint Multi-Hop Question Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing frameworks for multi-hop question generation lack high-dimensional semantic constraints. |
| Approach: | They propose a framework which incorporates constraints into every hop of a multi-hop question . they propose dimensional constraint dimension reduction and divide-and-conquer-based dynamic projection . |
| Outcome: | The proposed framework outperforms baseline models in capturing and integrating constraints . it reduces randomness with added constraints, leading to more focused questions and fewer hop errors . |
Can Indirect Prompt Injection Attacks Be Detected and Removed? (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have developed various detection mechanisms to protect against prompt injection attacks. |
| Approach: | They investigate the feasibility of detecting and removing indirect prompt injection attacks . they use two methods to evaluate their performance and train detection models . |
| Outcome: | The proposed method is based on a benchmark dataset and is available on github . it evaluates the performance of existing models and open-source detection models . |
Identifying Open Challenges in Language Identification (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on language identification has focused on cross-domain setups, but no systematic comparison is available. |
| Approach: | They propose to train an accurate multi-domain languageidentification model on 2,034 languages and analyze the remaining errors. |
| Outcome: | The proposed model performs well on 2,034 languages with training with 1,000 instances per language and a maximum input length of 100 characters. |
The Distracting Effect: Understanding Irrelevant Passages in RAG (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to detect and use hard distracting passages in RAG can cause problems . retrieved passages contain irrelevant but semantically related information that may mislead the LLM . |
| Approach: | They propose a method to identify and use hard distracting passages to improve RAG . they find that adding retrieved passages is found to ground the LLM response . |
| Outcome: | The proposed method achieves up to 7.5% increase in answering accuracy compared to fine-tuned datasets. |
Multilingual Encoder Knows more than You Realize: Shared Weights Pretraining for Extremely Low-Resource Languages (2025.acl-long)
Copied to clipboard
| Challenge: | XLM-R and mBART have advanced multilingualism in NLP, but low-resource languages such as Tibetan, Uyghur, Kazakh, and Mongolian are underserved. |
| Approach: | They propose a framework for adapting multilingual encoders to text generation in extremely low-resource languages by reusing the weights between the encoder and the decoder. |
| Outcome: | The proposed framework performs better on various downstream tasks even when compared with much larger models. |
Graphically Speaking: Unmasking Abuse in Social Media with Conversation Insights (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to detect abusive language often ignore conversational context, leading to inconsistent and sometimes inconclusive results. |
| Approach: | They propose a graph neural network approach that uses conversational context to model social media conversations as graphs, where nodes represent comments and edges capture reply structures. |
| Outcome: | The proposed model outperforms baseline and linear context-aware methods and achieves significant improvements in F1 scores. |
CodeTool: Enhancing Programmatic Tool Invocation of LLMs via Process Supervision (2025.acl-long)
Copied to clipboard
YifeiLu YifeiLu, Fanghua Ye, Jian Li, Qiang Gao, Cheng Liu, Haibo Luo, Nan Du, Xiaolong Li, Feiliang Ren
| Challenge: | Existing approaches to tool invocation are often unnecessarily long and require lengthy reasoning paths. |
| Approach: | They propose a framework for stepwise code generation that improves LLM tool invocation . they incorporate two distinct process rewards: the On-the-spot and the Latent Reward . |
| Outcome: | The proposed framework improves LLM tool invocation by leveraging the concise nature of code. |
RARE: Retrieval-Augmented Reasoning Enhancement for Large Language Models (2025.acl-long)
Copied to clipboard
Hieu Tran, Zonghai Yao, Zhichao Yang, Junda Wang, Yifan Zhang, Shuo Han, null Feiyun Ouyang, Hong Yu
| Challenge: | Existing work aims to improve reasoning accuracy and factual integrity across large language models for knowledge-intensive tasks such as medical and commonsense reasoning. |
| Approach: | They propose a versatile extension to the mutual reasoning framework (rStar) that enhances reasoning accuracy and factual integrity across large language models. |
| Outcome: | The proposed extension to the mutual reasoning framework improves reasoning accuracy and factual integrity across large language models for complex, knowledge-intensive tasks. |
Defense Against Prompt Injection Attack by Leveraging Attack Techniques (2025.acl-long)
Copied to clipboard
| Challenge: | Recent attacks leverage LLMs’ instruction-following abilities and their inabilities to distinguish instructions injected in the data content. |
| Approach: | They invert the intention of prompt injection methods to develop novel defense methods based on previous training-free attack methods by repeating the attack process with the original input instruction rather than the injected instruction. |
| Outcome: | The proposed methods outperform existing defense approaches, achieving state-of-the-art results. |
Acquisition and Application of Novel Knowledge in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for constructing new datasets rely on timestamps or simple templates that do not accurately reflect the real world. |
| Approach: | They propose a knowledge dataset construction approach that simulates biological evolution using knowledge graphs to generate synthetic entities with diverse attributes. |
| Outcome: | The proposed framework outperforms knowledge augmentation methods by 3.3%-38%. |
DNCASR: End-to-End Training for Speaker-Attributed ASR (2025.acl-long)
Copied to clipboard
| Challenge: | DNCASR is a novel end-to-end trainable system for speaker clustering and automatic speech recognition . it is designed to transcribe long multi-party meetings with speaker-attributed transcriptions . |
| Approach: | They propose a joint neural speaker clustering and automatic speech recognition system . DNCASR uses two separate encoders to independently encode speaker characteristics and local waveform information . |
| Outcome: | The proposed system outperforms a parallel system that does not have links between speaker and ASR decoders. |
Exploring Persona Sentiment Sensitivity in Personalized Dialogue Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Personalized dialogue systems have advanced with the integration of user-specific personas into large language models (LLMs). |
| Approach: | They propose a dialogue generation approach that explicitly accounts for persona polarity by combining a turn-based generation strategy with a profile ordering mechanism and sentiment-aware prompting. |
| Outcome: | The proposed approach accounts for persona polarity by combining a turn-based generation strategy with a profile ordering mechanism and sentiment-aware prompting. |
AntiLeakBench: Preventing Data Contamination by Automatically Constructing Benchmarks with Updated Real-World Knowledge (2025.acl-long)
Copied to clipboard
Xiaobao Wu, Liangming Pan, Yuxi Xie, Ruiwen Zhou, Shuai Zhao, Yubo Ma, Mingzhe Du, Rui Mao, Anh Tuan Luu, William Yang Wang
| Challenge: | Existing studies solve this challenge by updating benchmarks with newly collected data, but they fail to guarantee contamination-free evaluation as the newly collected knowledge may contain pre-existing knowledge. |
| Approach: | They propose an automated anti-leakage benchmarking framework that builds and updates benchmarks without human labor instead of using newly collected data. |
| Outcome: | The proposed framework significantly reduces the cost of benchmark maintenance to accommodate emerging LLMs. |
LLM-Guided Semantic-Aware Clustering for Topic Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that topic modeling is competitive compared to closed-source methods. |
| Approach: | They propose a semi-supervised topic modeling method that combines LLMs with clustering to improve topic generation and distribution. |
| Outcome: | The proposed method outperforms state-of-the-art methods that utilize GPT-4 on topic alignment and exhibits competitive performance compared to Neural Topic Models on topic quality. |
Hierarchical Bracketing Encodings for Dependency Parsing as Tagging (2025.acl-long)
Copied to clipboard
| Challenge: | Existing encodings for dependency parsing use suboptimal number of labels and a limited number of symbols. |
| Approach: | They propose a family of encodings for sequence labeling dependency parsing based on hierarchical bracketing . they propose an optimal hierarchically bracketing which minimizes the number of symbols used and encodes projective trees using only 12 distinct labels . |
| Outcome: | The proposed encodings yield competitive accuracy on a diverse set of treebanks. |
OASIS: Order-Augmented Strategy for Improved Code Search (2025.acl-long)
Copied to clipboard
Gao Zuchen, Zizheng Zhan, Xianming Li, Erxin Yu, Haotian Zhang, Chenbin Chenbin, Yuqun Zhang, Jing Li
| Challenge: | Code embeddings capture the semantic representations of code and are crucial for various code-related large language model (LLM) applications. |
| Approach: | They propose an order-augmented strategy for improved code search that leverages order-based similarity labels to capture subtle differences in similarity among negative pairs. |
| Outcome: | The proposed model outperforms state-of-the-art models focusing on major positive-negative differences. |
Can Large Language Models Detect Errors in Long Chain-of-Thought Reasoning? (2025.acl-long)
Copied to clipboard
Yancheng He, Shilong Li, Jiaheng Liu, Weixun Wang, Xingyuan Bu, Ge Zhang, Z.y. Peng, Zhaoxiang Zhang, Zhicheng Zheng, Wenbo Su, Bo Zheng
| Challenge: | Recent advances in o1-like models have generated long Chain-of-Thought reasoning steps to improve the reasoning abilities of existing Large Language Models (LLMs). |
| Approach: | They propose a DeltaBench to analyze the quality and effectiveness of o1-like models and measure their ability to detect errors in long COT reasoning. |
| Outcome: | The proposed model can detect errors in long COT reasoning. |
OmniAlign-V: Towards Enhanced Alignment of MLLMs with Human Preference (2025.acl-long)
Copied to clipboard
Xiangyu Zhao, Shengyuan Ding, Zicheng Zhang, Haian Huang, Maosongcao Maosongcao, Jiaqi Wang, Weiyun Wang, Xinyu Fang, Wenhai Wang, Guangtao Zhai, Hua Yang, Haodong Duan, Kai Chen
| Challenge: | Existing open-source multi-modal large language models (MLLMs) focus on enhancing foundational capabilities, leaving a significant gap in human preference alignment. |
| Approach: | They propose a dataset of 200K high-quality training samples featuring diverse images, complex questions, and varied response formats to improve MLLMs’ alignment with human preferences. |
| Outcome: | The proposed dataset of 200K high-quality training samples improves human preference alignment while maintaining or enhancing performance on standard VQA benchmarks. |
Tree-KG: An Expandable Knowledge Graph Construction Framework for Knowledge-intensive Domains (2025.acl-long)
Copied to clipboard
| Challenge: | Knowledge graphs are a useful tool for organizing complex data in knowledge-intensive domains. |
| Approach: | They propose an expandable framework that combines structured domain texts with advanced semantic techniques to create a tree-like graph from textbooks. |
| Outcome: | The proposed framework surpasses competing methods in the text-Annotated dataset with high scores on the Text-Annalytated data. |
Measuring Data Diversity for Instruction Tuning: A Systematic Analysis and A Reliable Metric (2025.acl-long)
Copied to clipboard
Yuming Yang, Yang Nan, Junjie Ye, Shihan Dou, Xiao Wang, Shuo Li, Huijie Lv, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Existing studies have explored various diversity-aware data selection methods to construct high-quality datasets and enhance model performance. |
| Approach: | They propose to use data diversity to measure instruction tuning of large language models. |
| Outcome: | The proposed diversity metric outperforms existing methods on simulated and real-world data and shows that it captures diversity variations and achieves a 0.97 correlation with instruction tuning. |
Micro-Act: Mitigate Knowledge Conflict in Question Answering via Actionable Self-Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to mitigate knowledge conflict by comparing two knowledge sources can overwhelm LLMs with extraneous or lengthy contexts. |
| Approach: | They propose a framework that decomposes knowledge into fine-grained comparisons . they propose 'Micro-Act' framework that allows for reasoning beyond the superficial context . |
| Outcome: | The proposed framework achieves significant increase in QA accuracy over state-of-the-art baselines on five benchmark datasets. |
Minimal Pair-Based Evaluation of Code-Switching (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods do not have wide language coverage, fail to account for the diverse range of CS phenomena, or do not scale. |
| Approach: | They propose to use minimal pairs of CS to estimate the extent to which large language models (LLMs) use code-switching in the same way as bilinguals. |
| Outcome: | The proposed model assigns higher probability to the naturally occurring CS sentence than to the variant for every language pair. |
DNASpeech: A Contextualized and Situated Text-to-Speech Dataset with Dialogues, Narratives and Actions (2025.acl-long)
Copied to clipboard
| Challenge: | Existing TTS datasets lack situated descriptive prompts aligned with speech data. |
| Approach: | They propose a contextualized and situated text-to-speech task to promote more accurate and customized speech generation using DNA prompts. |
| Outcome: | The proposed task promotes more accurate and customized speech generation using DNA prompts. |
LLaMA-Omni 2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | LLaMA-Omni 2 is a series of speech language models (SpeechLMs) based on large language models. |
| Approach: | They introduce a series of speech language models capable of real-time speech interaction . LLaMA-Omni 2 trains on 200K multi-turn speech dialogue samples . |
| Outcome: | The proposed speech language models surpass state-of-the-art models on spoken question answering and speech instruction. |
Error Comparison Optimization for Large Language Models on Aspect-Based Sentiment Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for aspect-based sentiment analysis (ABSA) only compare current predictions and labels on each sample, yet fail to perceive and understand its error outputs from different degrees. |
| Approach: | They propose a framework that can perceive and understand the degree of errors by learning from comparative error pairs. |
| Outcome: | The proposed framework exceeds baselines and achieves the desired performance. |
The AI Gap: How Socioeconomic Status Affects Language Technology Interactions (2025.acl-long)
Copied to clipboard
| Challenge: | Socioeconomic status (SES) fundamentally influences how people interact with technology, but it is limited by proxy metrics and synthetic data. |
| Approach: | They collect 6,482 prompts from previous interactions of 1,000 individuals from ‘diverse socioeconomic backgrounds’ about their use of language technologies and generative AI. |
| Outcome: | The findings show that higher SES groups have higher levels of abstraction, convey requests more concisely, and topics like ‘inclusivity’ and ‘travel’. |
Probing LLMs for Multilingual Discourse Generalization Through a Unified Label Set (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on discourse understanding is constrained by framework-dependent discourse representations. |
| Approach: | They examine whether large language models capture discourse knowledge that generalizes across languages and frameworks. |
| Outcome: | The proposed model can generalize discourse information across languages and frameworks. |
Crowdsource, Crawl, or Generate? Creating SEA-VL, a Multicultural Vision-Language Dataset for Southeast Asia (2025.acl-long)
Copied to clipboard
Samuel Cahyawijaya, Holy Lovenia, Joel Ruben Antony Moniz, Tack Hwa Wong, Mohammad Rifqi Farhansyah, Thant Thiri Maung, Frederikus Hudi, David Anugraha, Muhammad Ravi Shulthan Habibi, Muhammad Reza Qorib, Amit Agarwal, Joseph Marvin Imperial, Hitesh Laxmichand Patel, Vicky Feliren, Bahrul Ilmi Nasution, Manuel Antonio Rufino, Genta Indra Winata, Rian Adam Rajagede, Carlos Rafael Catalan, Mohamed Fazli Mohamed Imam, Priyaranjan Pattnayak, Salsabila Zahirah Pranida, Kevin Pratama, Yeshil Bangera, Adisai Na-Thalang, Patricia Nicole Monderin, Yueqi Song, Christian Simon, Lynnette Hui Xian Ng, Richardy Lobo Sapan, Taki Hasan Rafi, Bin Wang, null Supryadi, Kanyakorn Veerakanjana, Piyalitt Ittichaiwong, Matthew Theodore Roque, Karissa Vincentio, Takdanai Kreangphet, Phakphum Artkaew, Kadek Hendrawan Palgunadi, Yanzhi Yu, Rochana Prih Hastuti, William Nixon, Mithil Bangera, Adrian Xuan Wei Lim, Aye Hninn Khine, Hanif Muhammad Zhafran, Teddy Ferdinan, Audra Aurora Izzani, Ayushman Singh, Evan Evan, Jauza Akbar Krito, Michael Anugraha, Fenal Ashokbhai Ilasariya, Haochen Li, John Amadeo Daniswara, Filbert Aurelian Tjiaranata, Eryawan Presma Yulianrifat, Can Udomcharoenchaikit, Fadil Risdian Ansori, Mahardika Krisna Ihsani, Giang Nguyen, Anab Maulana Barik, Dan John Velasco, Rifo Ahmad Genadi, Saptarshi Saha, Chengwei Wei, Isaiah Edri W. Flores, Kenneth Chen Ko Han, Anjela Gail D. Santos, Wan Shen Lim, Kaung Si Phyo, Tim Santos, Meisyarah Dwiastuti, Jiayun Luo, Jan Christian Blaise Cruz, Ming Shan Hee, Ikhlasul Akmal Hanif, M.Alif Al Hakim, Muhammad Rizky Sya’ban, Kun Kerdthaisong, Lester James Validad Miranda, Fajri Koto, Tirana Noor Fatyanosa, Alham Fikri Aji, Jostin Jerico Rosal, Jun Kevin, Robert Wijaya, Onno P. Kampman, Ruochen Zhang, Börje F. Karlsson, Peerat Limkonchotiwat
| Challenge: | Southeast Asia is underrepresented in vision-language research . SEA-VL is an open-source initiative dedicated to developing culturally relevant datasets for SEA languages. |
| Approach: | They propose to use crowdsourced, automated image crawling and synthetic image generation to develop culturally relevant datasets for SEA languages. |
| Outcome: | The proposed datasets capture SEA cultural nuances and contexts better than existing datasets. |
Soundwave: Less is More for Speech-Text Alignment in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing end-to-end speech large language models rely on large-scale annotated data for training, while data-efficient training has not been discussed in depth. |
| Approach: | They propose a training strategy and a novel architecture to address representation space gap and sequence length inconsistency in speech and text. |
| Outcome: | The proposed model outperforms other advanced speech LLMs in speech translation and AIR-Bench speech tasks with only a fraction of the training data. |
RoToR: Towards More Reliable Responses for Order-Invariant Inputs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing solutions to positional bias in listwise inputs are limited on practical problems . e.g., lost-in-the-middle problem is a common problem for listwise models . |
| Approach: | They propose a zero-shot order-invariant LM for order- invariant inputs with minimal modifications of positional IDs and Selective Routing for listwise tasks. |
| Outcome: | The proposed framework can handle order-invariant and sensitive inputs in listwise tasks. |
Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation (2025.acl-long)
Copied to clipboard
Shivalika Singh, Angelika Romanou, Clémentine Fourrier, David Ifeoluwa Adelani, Jian Gang Ngui, Daniel Vila-Suero, Peerat Limkonchotiwat, Kelly Marchisio, Wei Qi Leong, Yosephine Susanto, Raymond Ng, Shayne Longpre, Sebastian Ruder, Wei-Yin Ko, Antoine Bosselut, Alice Oh, Andre Martins, Leshem Choshen, Daphne Ippolito, Enzo Ferrante, Marzieh Fadaee, Beyza Ermis, Sara Hooker
| Challenge: | Reliable multilingual evaluation is difficult and culturally appropriate evaluation is even harder to achieve. |
| Approach: | They propose a multilingual evaluation framework that aims to mitigate these biases by improving translations and annotation practices. |
| Outcome: | The proposed framework improves translation quality and cultural coverage and is culturally sensitive and culturally agnostic. |
Improving Dialogue Discourse Parsing through Discourse-aware Utterance Clarification (2025.acl-long)
Copied to clipboard
| Challenge: | Extensive experiments on the STAC and Molweni datasets demonstrate that our approach effectively resolves ambiguities and significantly outperforms the state-of-the-art (SOTA) baselines. |
| Approach: | They propose a Discourse-aware Clarification Module (DCM) that generates clarifications for the parser through systematic clarification type reasoning and discourse goal reasoning. |
| Outcome: | Extensive experiments on the STAC and Molweni datasets demonstrate that the proposed module significantly outperforms the state-of-the-art (SOTA) framework. |
ImPart: Importance-Aware Delta-Sparsification for Improved Model Compression and Merging in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent approaches to reduce resource requirements for task-specific large language models have been developed. |
| Approach: | They propose a delta compression approach that optimizes for importance of a model . they use SVD to dynamically adjust the sparsity ratios of different vectors based on their importance . |
| Outcome: | The proposed approach achieves state-of-the-art in retaining task-specific knowledge even at high sparsity ratios. |
Words of Warmth: Trust and Sociability Norms for over 26k English Words (2025.acl-long)
Copied to clipboard
| Challenge: | Social psychologists have shown that Warmth (W) and Competence (C) are the primary dimensions along which we assess other people and groups. |
| Approach: | They propose a repository of word–warmth and word–trust associations for over 26k English words. |
| Outcome: | The proposed lexicon enables bias and stereotype research through case studies on target entities. |
BehaviorBox: Automated Discovery of Fine-Grained Performance Differences Between Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for evaluating language models are brittle, corpus-level perplexities are vague, and the choice of benchmarks is endless. |
| Approach: | They propose a method that uses contextual embeddings to find fine-grained features of text where one model outperforms another. |
| Outcome: | The proposed method extracts features that demonstrate differences with respect to ease of generation between two language models. |
HAF-RM: A Hybrid Alignment Framework for Reward Model Training (2025.acl-long)
Copied to clipboard
Shujun Liu, Xiaoyu Shen, Yuhang Lai, Siyuan Wang, Shengbin Yue, Zengfeng Huang, Xuanjing Huang, Zhongyu Wei
| Challenge: | Recent studies have focused on enhancing reward models through data improvements, following the conventional training framework for reward models that directly optimizes the predicted rewards. |
| Approach: | They propose a hybrid alignment framework **HAF-RM** that incorporates additional constraint on token-level policy probabilities in addition to the reward score. |
| Outcome: | The proposed framework can supervise the internal preference model at the token level and optimize the mapping layer of the reward model at sequence level. |
CULEMO: Cultural Lenses on Emotion - Benchmarking LLMs for Cross-Cultural Emotion Understanding (2025.acl-long)
Copied to clipboard
Tadesse Destaw Belay, Ahmed Haj Ahmed, Alvin C Grissom Ii, Iqra Ameer, Grigori Sidorov, Olga Kolesnikova, Seid Muhie Yimam
| Challenge: | Existing emotion benchmarks rely on keyword-based emotion recognition, overlooking cultural dimensions required for emotion understanding. |
| Approach: | They propose a benchmark to evaluate culturally-aware emotion prediction across six languages. |
| Outcome: | The proposed benchmark evaluates state-of-the-art LLMs on culture-aware emotion prediction and sentiment analysis tasks. |
DiffPO: Diffusion-styled Preference Optimization for Inference Time Alignment of Large Language Models (2025.acl-long)
Copied to clipboard
Ruizhe Chen, Wenhao Chai, Zhifei Yang, Xiaotian Zhang, Ziyang Wang, Tony Quek, Joey Tianyi Zhou, Soujanya Poria, Zuozhu Liu
| Challenge: | Inference-time alignment approaches still face limitations due to policy-specific value functions and latency during the inference phase. |
| Approach: | They propose an efficient and policy-agnostic preference optimization method that avoids time latency associated with token generation. |
| Outcome: | The proposed method achieves a favorable trade-off between alignment quality and inference-time latency. |
MemeQA: Holistic Evaluation for Meme Understanding (2025.acl-long)
Copied to clipboard
Khoi P. N. Nguyen, Terrence Li, Derek Lou Zhou, Gabriel Xiong, Pranav Balu, Nandhan Alahari, Alan Huang, Tanush Chauhan, Harshavardhan Bala, Emre Guzelordu, Affan Kashfi, Aaron Xu, Suyesh Shrestha, Megan Vu, Jerry Wang, Vincent Ng
| Challenge: | Existing benchmarks for meme understanding only concern narrow aspects of meme semantics. |
| Approach: | They propose to use multiple-choice questions to evaluate meme comprehension . they use a dataset of over 9,000 multiple-question questions to assess meme comprehension. |
| Outcome: | The proposed model outperforms existing models on meme comprehension . the model makes many errors on memes where proper understanding requires going beyond sentiment . |
LoGU: Long-form Generation with Uncertainty Expressions (2025.acl-long)
Copied to clipboard
Ruihan Yang, Caiqi Zhang, Zhisong Zhang, Xinting Huang, Sen Yang, Nigel Collier, Dong Yu, Deqing Yang
| Challenge: | Large Language Models (LLMs) generate factually incorrect content, i.e., hallucinations, despite impressive performance. |
| Approach: | They propose a framework to enable models to express uncertainty when unsure . they propose atomic claims to refine uncertainty and refine it using supervised fine-tuning and direct preference optimization to enhance uncertainty expression. |
| Outcome: | The proposed framework significantly improves accuracy, reduces hallucinations, and maintains comprehensiveness of responses. |
KiRAG: Knowledge-Driven Iterative Retriever for Enhancing Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Iterative retrieval-augmented generation models are difficult to use for multihop question answering (QA) . their retrieval processes can be disrupted by irrelevant documents or factually inaccurate chain-of-thoughts . |
| Approach: | They propose a knowledge-driven iterative retriever model that decomposes documents into knowledge triples and performs iterativ retrieval with these triples to enable a factually reliable retrieval process. |
| Outcome: | The proposed model outperforms existing iRAG models with an average improvement of 9.40% in R@3 and 5.14% in F1 on multi-hop QA datasets. |
Enhancing Lexicon-Based Text Embeddings with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent large language models (LLMs) have demonstrated exceptional performance on general-purpose text embedding tasks. |
| Approach: | They introduce the first lexicon-based embeddings that consolidates the vocabulary space through token embeddation clustering to handle the issue of token redundancy in LLM vocabularies. |
| Outcome: | The proposed model outperforms dense embeddings on the Massive Text Embedding Benchmark (MTEB) it also supports efficient dimension pruning without any specialized objectives like Matryoshka Representation Learning. |
CoCoLex: Confidence-guided Copy-based Decoding for Grounded Legal Text Generation (2025.acl-long)
Copied to clipboard
Santosh T.y.s.s, Youssef Tarek Elkhayat, Oana Ichim, Pranav Shetty, Dongsheng Wang, Zhiqiang Ma, Armineh Nourbakhsh, Xiaomo Liu
| Challenge: | LLMs can provide key benefits to the Legal domain, but their adoption has been hindered by their tendency to generate unfaithful, ungrounded, or hallucinatory outputs. |
| Approach: | They propose a Confidence-guided copy-based decoding strategy that dynamically interpolates the model produced vocabulary distribution with a distribution derived based on copying from the context. |
| Outcome: | The proposed method outperforms existing context-aware decoding methods on five legal benchmarks. |
Beyond N-Grams: Rethinking Evaluation Metrics and Strategies for Multilingual Abstractive Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | n-gram-based metrics are considered indicative (even if imperfect) of human evaluation for English, but their suitability for other languages remains unclear. |
| Approach: | They systematically assess evaluation metrics for generation for languages and tasks using n-gram-based and neural-based metrics. |
| Outcome: | The proposed evaluation suite is based on eight languages from four typological families and shows that it is sensitivity to the language type at hand. |
CC-Tuning: A Cross-Lingual Connection Mechanism for Improving Joint Multilingual Supervised Fine-Tuning (2025.acl-long)
Copied to clipboard
Yangfan Ye, Xiaocheng Feng, Zekun Yuan, Xiachong Feng, Libo Qin, Lei Huang, Weitao Ma, Yichong Huang, Zhirui Zhang, Yunfei Lu, Xiaohui Yan, Duyu Tang, Dandan Tu, Bing Qin
| Challenge: | Existing fine-tuning approaches that focus on English-centric training corpora often introduce implicit cross-lingual alignment, overlooking the potential for more profound, latent-level cross-linguistic interactions. |
| Approach: | They propose a multilingual fine-tuning paradigm that explicitly establishes a cross-lingual connection mechanism at the latent level. |
| Outcome: | The proposed model outperforms vanilla SFT and offers a strong latent-level alternative to data-level augmentation methods. |
SConU: Selective Conformal Uncertainty in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing frameworks fail to identify outliers that violate the exchangeability assumption, leading to unbounded miscoverage rates and unactionable prediction sets. |
| Approach: | They propose a method that implements significance tests to determine whether a given sample deviates from the uncertainty distribution of the calibration set. |
| Outcome: | The proposed approach facilitates rigorous management of miscoverage rates across single-domain and interdisciplinary contexts, and enhances the efficiency of predictions. |
MegaPairs: Massive Data Synthesis for Universal Multimodal Retrieval (2025.acl-long)
Copied to clipboard
Junjie Zhou, Yongping Xiong, Zheng Liu, Ze Liu, Shitao Xiao, Yueze Wang, Bo Zhao, Chen Jason Zhang, Defu Lian
| Challenge: | despite the growing demand for multimodal retrieval, there is a lack of training data. |
| Approach: | They propose a data synthesis method that leverages vision language models and open-domain images to generate high-quality data. |
| Outcome: | The proposed method outperforms baseline models on 70 more datasets and can scale up. |
When GPT Spills the Tea: Comprehensive Assessment of Knowledge File Leakage in GPTs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that adversarial prompts can induce GPTs to leak knowledge file content. |
| Approach: | They propose a workflow inspired by Data Security Posture Management to identify five leakage vectors for knowledge file leakage using 651,022 GPT metadata and 11,820 flows. |
| Outcome: | The proposed workflow analyzes 651,022 GPT metadata, 11,820 flows, and 1,466 responses to identify five leakage vectors: metadata, GPT initialization, retrieval, sandboxed execution environments, and prompts. |
UniCodec: Unified Audio Codec with Single Domain-Adaptive Codebook (2025.acl-long)
Copied to clipboard
Yidi Jiang, Qian Chen, Shengpeng Ji, Yu Xi, Wen Wang, Chong Zhang, Xianghu Yue, ShiLiang Zhang, Haizhou Li
| Challenge: | Existing neural audio codecs are not capable of handling multi-domain audio data . et al., 2023) integrate speech modality with text-based large language models . |
| Approach: | They propose a unified audio codec with a single codebook to support multi-domain audio data . they propose combining a mix-of-experts strategy and a partitioned domain-adaptive codebook method . |
| Outcome: | The proposed codec outperforms existing codecs on acoustic and semantic representation capabilities. |
KERL: Knowledge-Enhanced Personalized Recipe Recommendation using Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models and the abundance of food data have led to studies to improve food understanding using LLMs. |
| Approach: | They propose a unified system that leverages food KGs and LLMs to provide personalized food recommendations and generate recipes with associated micro-nutritional information. |
| Outcome: | The proposed system outperforms existing methods in food recommendation, recipe generation and nutritional analysis. |
Multilingual Arbitration: Optimizing Data Pools to Accelerate Multilingual Progress (2025.acl-long)
Copied to clipboard
| Challenge: | Synthetic data generation relies on a single oracle teacher model, which can lead to model collapse and bias propagation. |
| Approach: | They propose a multilingual arbitration approach that exploits performance variations among multiple models for each language. |
| Outcome: | The proposed approach surpasses single-teacher distillation with 80% win rates over proprietary and open-weight models with the largest improvements in low-resource languages. |
Controlled Low-Rank Adaptation with Subspace Regularization for Continued Training on Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to mitigate catastrophic forgetting can be broadly categorized into data-based, architecture-based and learning-based methods. |
| Approach: | They propose a subspace regularization method on LoRA structure that imposes constraints on direction of updating matrix’s null space. |
| Outcome: | The proposed method reduces scale of output change while introducing minimal constraint on model capacity. |
Chinese SimpleQA: A Chinese Factuality Evaluation for Large Language Models (2025.acl-long)
Copied to clipboard
Yancheng He, Shilong Li, Jiaheng Liu, Yingshui Tan, Weixun Wang, Hui Huang, Xingyuan Bu, Hangyu Guo, Chengwei Hu, Boren Zheng, Zhuoran Lin, Dekai Sun, Zhicheng Zheng, Wenbo Su, Bo Zheng
| Challenge: | Current frontier models sometimes generate false outputs or answers that are not substantiated by evidence. |
| Approach: | They propose Chinese SimpleQA, a Chinese benchmark to evaluate LLMs' factuality . they focus on Chinese language over 6 major topics with 99 diverse subtopics . |
| Outcome: | The Chinese SimpleQA benchmark evaluates the factuality ability of LLMs . the questions and answers are short and easy-to-evaluate . |
PVP: An Image Dataset for Personalized Visual Persuasion with Persuasion Strategies, Viewer Characteristics, and Persuasiveness Ratings (2025.acl-long)
Copied to clipboard
| Challenge: | Visual persuasion uses visual elements to influence cognition and behaviors . lack of comprehensive data sets connect persuasiveness of images with personal information . |
| Approach: | They propose to use a dataset to connect persuasiveness with personal information . they find psychological characteristics enhance the generation and evaluation of persuasive images . |
| Outcome: | The proposed dataset provides persuasiveness scores of images evaluated by human annotators along with demographic and psychological characteristics. |
Any Information Is Just Worth One Single Screenshot: Unifying Search With Visualized Information Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing multimodal retrieval models are lacking in visual representations of multimodal data. |
| Approach: | They propose a visualized information retrieval paradigm where multimodal information is represented by a unified visual format called Screenshots for various retrieval applications. |
| Outcome: | The proposed model is based on a large dataset of screenshots from diverse sources . it is compared with existing models and lays a solid foundation for the new model . |
Tunable LLM-based Proactive Recommendation Agent (2025.acl-long)
Copied to clipboard
| Challenge: | Current methods focus on catering to existing user interests, leading to polarized recommendation distributions. |
| Approach: | They propose an LLM-based Actor-Critic Agent framework to cultivate latent interests through multi-step recommendations. |
| Outcome: | The proposed framework optimizes long-term rewards and dynamically evolves with user feedback. |
AgentRM: Enhancing Agent Generalization with Reward Modeling (2025.acl-long)
Copied to clipboard
Yu Xia, Jingru Fan, Weize Chen, Siyu Yan, Xin Cong, Zhong Zhang, Yaxi Lu, Yankai Lin, Zhiyuan Liu, Maosong Sun
| Challenge: | Existing LLM-based agents have strong performance on held-in tasks, but their generalizability to unseen tasks remains poor. |
| Approach: | They propose a reward-based generalizable reward model to guide the policy model for effective test-time search. |
| Outcome: | The proposed agentRM outperforms existing agents on held-in tasks by 8.8 points on average. |
From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to align large language models with human preferences suffer from inconsistent scoring and suboptimal alignment. |
| Approach: | They propose a dual-consistency framework that aligns partial sequences with human preferences. |
| Outcome: | The proposed framework significantly reduces granularity discrepancies and improves GPT-4 evaluation scores. |
Segment-Based Attention Masking for GPTs (2025.acl-long)
Copied to clipboard
| Challenge: | Causal masking is a fundamental component in Generative Pre-Trained Transformers (GPT) models, playing a crucial role during training. |
| Approach: | They propose to apply causal masking to all input tokens step-by-step, mimicking the generation process. |
| Outcome: | The proposed model can process the entire user prompt at once, but it is applied to all input tokens step-by-step, mimicking the generation process. |
Cramming 1568 Tokens into a Single Vector and Back Again: Exploring the Limits of Embedding Space Capacity (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work addresses problem of compression of tokens into shorter sequence of real-valued vectors . attainable lossless compression ratio is typically not higher than x10 . |
| Approach: | They propose to compress a sequence of tokens into a shorter sequence of real-valued vectors to be used as inputs instead of token embeddings or key-value cache. |
| Outcome: | The proposed algorithms reduce the amount of compute in existing language models rather than minimizing number of bits needed to store text. |
Bi-Tuning with Collaborative Information for Controllable LLM-based Sequential Recommendation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to optimize sequential recommendation systems rely on item ID sequences, but they lack collaborative knowledge and limited controllability. |
| Approach: | They propose a simple bi-tuning framework with collaborative information for controllable Large Language Model-based Sequential Recommendation (Laser) they incorporate learnable virtual tokens at prefix and suffix of input text to adapt LLMs with collaborative knowledge . |
| Outcome: | The proposed framework outperforms state-of-the-art recommendations on real-world datasets. |
A Modular Approach for Clinical SLMs Driven by Synthetic Data with Pre-Instruction Tuning, Model Merging, and Clinical-Tasks Alignment (2025.acl-long)
Copied to clipboard
Jean-Philippe Corbeil, Amin Dada, Jean-Michel Attendu, Asma Ben Abacha, Alessandro Sordoni, Lucas Caccia, Francois Beaulieu, Thomas Lin, Jens Kleesiek, Paul Vozila
| Challenge: | Large language models such as GPT-4 have limited their deployment in clinical settings . a novel framework for adapting SLMs into high-performing clinical models is needed . |
| Approach: | They propose a framework for adapting large language models into high-performing clinical models . they pre-instruct experts on relevant medical and clinical corpora and model merging . |
| Outcome: | The proposed framework outperforms the existing model on the CLUE+ benchmark on medical entities and radiology reports. |
DIVE into MoE: Diversity-Enhanced Reconstruction of Large Language Models from Dense into Mixture-of-Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for reconstruction of large language models overlook diversity among experts, leading to potential redundancy. |
| Approach: | They propose a pruning-based expert reconstruction method that prunes a specific LLM and retrains it on routers, experts and normalization modules. |
| Outcome: | The proposed method outperforms pruning and MoE reconstruction methods on Llama-style models with open-source training corpora. |
DAC: A Dynamic Attention-aware Approach for Task-Agnostic Prompt Compression (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods rely on information entropy as the metric to compress lexical units, but ignore attention-critical tokens and information . recent advent of In-Context Learning (ICL), Chain-of-Thought (CoT), and Retrieval Augmented Generation (RAG) technologies has significantly invigorated the landscape of applications based on Large Language Models (LLMs). |
| Approach: | They propose a dynamic attention-aware approach to task-agnostic prompt compression . they integrate entropy and attention information to achieve fine-grained prompt compression. |
| Outcome: | Experiments show that the proposed approach improves across tasks and LLMs. |
Computation Mechanism Behind LLM Position Generalization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have explored how LLMs handle positional relevance, but how they handle it remains unexplored. |
| Approach: | They propose to enforce certain computational mechanisms to allow for the tolerance in position perturbations in large language models (LLMs) they also find a pattern in intermediate features that allows this effect to be observed . |
| Outcome: | The proposed models can understand text with position perturbations and generalize to longer sequences than those seen during training with the latest techniques. |
IPO: Your Language Model is Secretly a Preference Classifier (2025.acl-long)
Copied to clipboard
| Challenge: | Reinforcement learning from human feedback (RLHF) is the primary method for aligning large language models with human preferences, but it often incurs significant computational and financial costs due to its reliance on training external reward models or human-labeled preferences. |
| Approach: | They propose an alternative approach that leverages generative LLMs as preference classifiers to reduce the dependence on external reward models or human-labeled preferences. |
| Outcome: | The proposed approach reduces the dependence on external reward models or human-labeled preferences by using generative LLMs as preference classifiers. |
Reversal of Thought: Enhancing Large Language Models with Preference-Guided Reverse Reasoning Warm-up (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve LLMs’ logical capabilities involve traceable or verifiable logical sequences that generate more reliable responses yet increase computational costs, or introduce rigid logic template rules, reducing flexibility. |
| Approach: | They propose a plug-and-play reasoning framework that enhances LLMs' logical reasoning abilities during the warm-up phase prior to batch inference. |
| Outcome: | The proposed framework surpasses baselines in both reasoning accuracy and efficiency. |
Déjà Vu? Decoding Repeated Reading from Eye Movements (2025.acl-long)
Copied to clipboard
| Challenge: | In many daily situations we read the same text more than once. |
| Approach: | They propose a strategy for enhancing feature-based and neural models by generating machine generated eye movements from a cognitive model. |
| Outcome: | The proposed model improves on the previous model and enables better characterization of the role of memory in repeated reading. |
LLMs can be easily Confused by Instructional Distractions (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) show exceptional skill in instruction following tasks, but can become vulnerable when they are required to disregard instructions. |
| Approach: | They propose a benchmark to assess LLMs' performance under instructional distraction. |
| Outcome: | The proposed benchmark categorizes real-world instances of instructional distraction and evaluates LLMs across four instruction tasks: proofreading, rewriting, translation, and style transfer—alongside five input tasks: reasoning, code generation, mathematical reasoning, bias detection, and question answering. |
PlanGenLLMs: A Modern Survey of LLM Planning Capabilities (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have focused on developing LLMs to automate complex planning tasks. |
| Approach: | They propose to provide a comprehensive overview of current LLM planners to fill this gap . they examine performance criteria including completeness, executability, optimality, representation, generalization, and efficiency . |
| Outcome: | The proposed survey examines performance criteria for LLM planners and highlights their strengths and weaknesses. |
IAM: Efficient Inference through Attention Mapping between Different-scale LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are a challenge due to their internal reasoning processes. |
| Approach: | They propose an algorithm that can optimize attention matrices by performing attention mapping between small and large LLMs. |
| Outcome: | The proposed framework can reduce KV cache usage by 22.1% and accelerate prefill by 15% without sacrificing performance. |
nvAgent: Automated Data Visualization from Natural Language via Collaborative Agent Workflow (2025.acl-long)
Copied to clipboard
| Challenge: | *Natural Language to Visualization (NL2Vis) seeks to transform natural-language descriptions into visual representations of given tables. |
| Approach: | They propose a collaborative agent workflow for NL2Vis that incorporates three agents . the model is called **nvAgent** and comprises a processor agent for database processing and context filtering, a composer agent for planning visualization generation and a validator agent for code translation and output verification. |
| Outcome: | The proposed workflow surpasses state-of-the-art models on the VisEval benchmark. |
ZIPA: A family of efficient models for multilingual phone recognition (2025.acl-long)
Copied to clipboard
| Challenge: | IPA transcriptions capture major articulatory contrasts in speech sounds, including the voicing status, place of articulation, manner of voicing, and tongue positions. |
| Approach: | They present ZIPA, a family of efficient speech models that advances the state-of-the-art performance of crosslinguistic phone recognition. |
| Outcome: | The proposed model outperforms existing phone recognition systems on 17,000+ hours of normalized phone transcriptions and a novel evaluation set capturing unseen languages and sociophonetic variation. |
GRACE: A Granular Benchmark for Evaluating Model Calibration against Human Calibration (2025.acl-long)
Copied to clipboard
| Challenge: | Language models are often miscalibrated, leading to confidently incorrect answers. |
| Approach: | They propose a benchmark for language model calibration that incorporates comparison with human calibration. |
| Outcome: | The proposed metric analyzes model calibration errors and identifies types of miscalibration that differ from human behavior. |
Dynamic Evaluation with Cognitive Reasoning for Multi-turn Safety of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing safety evaluation methods rely on static assessments that use fixed harmful prompts or predefined prefixes as jailbreak templates. |
| Approach: | They propose a dynamic evaluation framework for multi-turn safety assessment of LLMs based on cognitive theories to simulate real chatting process and scenario simulation and strategy decision to guide dynamic generation. |
| Outcome: | The proposed framework has been applied to evaluate the safety of widely used LLMs. |
From Tools to Teammates: Evaluating LLMs in Multi-Session Coding Interactions (2025.acl-long)
Copied to clipboard
Nathanaël Carraz Rakotonirina, Mohammed Hamdy, Jon Ander Campos, Lucas Weber, Alberto Testoni, Marzieh Fadaee, Sandro Pezzelle, Marco Del Tredici
| Challenge: | Large Language Models excel at solving individual problems in isolation, but are they able to effectively collaborate over long-term interactions? |
| Approach: | They propose to use a multi-session dataset to test LLMs' ability to track and execute simple coding instructions amid irrelevant information, simulating a realistic setting. |
| Outcome: | The proposed model performs poorly when instructions are spread across sessions, suggesting that they are not able to integrate information over long interactions. |
Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints (2025.acl-long)
Copied to clipboard
| Challenge: | Jailbreaking attacks can effectively induce unsafe behaviors in Large Language Models (LLMs). |
| Approach: | They propose a conceptual framework to elucidate transferability of gradient-based jailbreaking methods . they identify superfluous constraints as significant barriers to improved transferability . |
| Outcome: | The proposed method increases the overall transfer attack success rate (T-ASR) across target and source models with varying safety levels from 18.4% to 50.3% while improving stability and controllability of jailbreak behaviors. |
Multilingual Text-to-Image Generation Magnifies Gender Stereotypes (2025.acl-long)
Copied to clipboard
Felix Friedrich, Katharina Hämmerl, Patrick Schramowski, Manuel Brack, Jindřich Libovický, Alexander Fraser, Kristian Kersting
| Challenge: | Text-to-image (T2I) generation models have great results in image quality, flexibility, and text alignment, but they suffer from substantial gender bias. |
| Approach: | They propose a benchmark to study gender bias in multilingual T2I models . they use multilingual prompts to account for grammatical differences influencing gender . |
| Outcome: | The proposed benchmark shows strong gender biases and language-specific differences across models. |
Adversarial Alignment with Anchor Dragging Drift (A3D2): Multimodal Domain Adaptation with Partially Shifted Modalities (2025.acl-long)
Copied to clipboard
| Challenge: | Domain adaptation is underexplored in multimodal learning environments due to expensive data collection and annotation. |
| Approach: | They propose a bi-alignment scheme to perform drift-drift and anchor-driving matching with partially shifting anchors. |
| Outcome: | The proposed approach achieves superior performance compared with state-of-the-art approaches. |
A Reality Check on Context Utilisation for Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
Lovisa Hagström, Sara Vera Marjanovic, Haeun Yu, Arnav Arora, Christina Lioma, Maria Maistro, Pepa Atanasova, Isabelle Augenstein
| Challenge: | Existing studies on LM context utilisation of retrieved information have focused on synthetic text. |
| Approach: | They propose a dataset of unreliable, insufficient and difficult-to-understand contexts with real-world queries and contexts manually annotated for stance to compare them to synthetic datasets. |
| Outcome: | The proposed model outperforms synthetic datasets and exaggerates rare context characteristics, leading to inflated context utilisation results. |
CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining (2025.acl-long)
Copied to clipboard
| Challenge: | Argument Mining (AM) involves the automatic identification of argument structure in natural language. |
| Approach: | They propose an approach that captures local and global coherence to identify argument structures by modeling macro-structure. |
| Outcome: | The proposed approach shows superior performance on heterogeneous datasets and on unseen datasets. |
Safer or Luckier? LLMs as Safety Evaluators Are Not Robust to Artifacts (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly employed as automated evaluators to assess the safety of generated content. |
| Approach: | They evaluate 11 LLM judge models across critical safety domains . apologetic language artifacts alone can skew evaluator preferences by up to 98% . |
| Outcome: | The findings show that apologetic language artifacts alone can skew evaluator preferences by up to 98% . |
Text-to-ES Bench: A Comprehensive Benchmark for Converting Natural Language to Elasticsearch Query (2025.acl-long)
Copied to clipboard
DonggeXue DonggeXue, Zhili Pu, Zhentao Xia, Hongli Sun, Ruihui Hou, Guangya Yu, Yupian Lin, Yongqi Fan, Jingping Liu, Tong Ruan
| Challenge: | Recent research on text-to-Query has explored using large language models to convert user query intent to executable code. |
| Approach: | They propose a novel semantic parsing task that leverages large language models to generate domain-specific language and post-processing code to support multi-index Elasticsearch queries. |
| Outcome: | The proposed model outperforms DeepSeek-R1 on the large Elasticsearch Dataset (LED) and BirdES datasets. |
AlignDistil: Token-Level Language Model Alignment as Adaptive Policy Distillation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for LLM alignment optimize tokens using a sparse, response-level reward or preference annotation. |
| Approach: | They propose an RLHF-equivalent distillation method for token-level reward optimization that incorporates the reward learned by DPO into the RLHG objective and builds a token-based teacher distribution. |
| Outcome: | The proposed method bridges the accuracy gap between the reward from the DPO model and the pure reward model by building a contrastive DPO reward with a normal and a reverse DPO. |
DARS: Dynamic Action Re-Sampling to Enhance Coding Agent Performance by Adaptive Tree Traversal (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to developing LLM-powered coding agents struggle with sub-optimal decision-making. |
| Approach: | They propose a novel inference time compute scaling approach that recovers from sub-optimal decisions by branching out a trajectory at certain key decision points by taking an alternative action given the history of the trajectory and execution feedback of the previous attempt. |
| Outcome: | The proposed approach achieves a pass@1 rate of 47% on the SWE-Bench Lite benchmark, outperforming state-of-the-art (SOTA) open-source frameworks. |
Steering off Course: Reliability Challenges in Steering Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Prior studies have evaluated a few steering methods for language models, leaving gaps in understanding their robustness. |
| Approach: | They examine three steering methods for language models to examine their reliability . they use function vectors, task vectors and DoLa to steer models toward desirable outputs . |
| Outcome: | The proposed methods show that they are not robust enough to handle large models with large parameters. |
Impartial Multi-task Representation Learning via Variance-invariant Probabilistic Decoding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods focus on balancing loss or gradients but fail to address this issue due to the representation discrepancy in latent space. |
| Approach: | They propose a framework that harmonizes representation spaces across tasks to ensure impartial learning by harmonizing representation spaces. |
| Outcome: | The proposed framework outperforms 12 representative methods under the same multi-task settings, especially in heterogeneous task combinations and data-constrained scenarios. |
If Eleanor Rigby Had Met ChatGPT: A Study on Loneliness in a Post-LLM World (2025.acl-long)
Copied to clipboard
| Challenge: | Loneliness is a global health concern and is prevalent worldwide . |
| Approach: | They analysed user interactions with ChatGPT outside of its marketed use as a task-oriented assistant and found that LLMs are more prevalent and riskier than LLM-based services . |
| Outcome: | The proposed models modify the LLMs to respond to loneliness and provide better engagement in conversations. |
Integrating Audio, Visual, and Semantic Information for Enhanced Multimodal Speaker Diarization on Multi-party Conversation (2025.acl-long)
Copied to clipboard
Luyao Cheng, Hui Wang, Chong Deng, Siqi Zheng, Yafeng Chen, Rongjie Huang, Qinglin Zhang, Qian Chen, Xihao Li, Wen Wang
| Challenge: | Mainstream speaker diarization systems rely only on acoustic information, making it challenging in complex aural environments. |
| Approach: | They propose a multimodal approach that integrates audio, visual, and semantic cues to enhance speaker diarization. |
| Outcome: | The proposed approach outperforms state-of-the-art methods on multi-party conversations . it integrates audio-visual-semantic cues into the clustering process for acoustic speaker embeddings . |
Vulnerability of LLMs to Vertically Aligned Text Manipulations (2025.acl-long)
Copied to clipboard
| Challenge: | Recent research shows that vertical text input significantly degrades the accuracy of large language models (LLMs) in text classification tasks. |
| Approach: | They investigate the impact of vertical text input on the performance of LLMs . they find that chain of thought reasoning does not help LLM recognize vertical input . |
| Outcome: | The proposed model can significantly mislead models, posing a risk of bypassing detection in real-world scenarios involving harmful or sensitive information. |
AutoMixer: Checkpoint Artifacts as Automatic Data Mixers (2025.acl-long)
Copied to clipboard
| Challenge: | In language model training, it is difficult to obtain the right data mixtures for various tasks as the relationship between data and tasks is difficult. |
| Approach: | They propose to identify checkpoint models based on their respective capabilities and leverage them as data mixers by using their aggregated first-order influence approximation over source data. |
| Outcome: | The proposed framework shows significant improvements on eight reasoning benchmarks, with accuracy increases of up to 1.93%. |
Generalized Attention Flow: Feature Attribution for Transformer Models via Maximum Flow (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to feature attributions rely on attention weights and attention weightings. |
| Approach: | They propose a feature attribution method that replaces attention weights with the generalized Information Tensor to enhance the performance of Transformer-based models. |
| Outcome: | The proposed method outperforms state-of-the-art feature attribution methods on sequence classification tasks and provides a more reliable interpretation of Transformer model outputs. |
Beyond Prompting: An Efficient Embedding Framework for Open-Domain Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have recently pushed open-domain question answering (ODQA) to new heights. |
| Approach: | They propose an embedding-level framework that enhances both the retriever and the reader by reordering query representations via lightweight linear layers under an unsupervised contrastive learning objective. |
| Outcome: | The proposed framework outperforms baselines in accuracy and efficiency across three open-source LLMs, three retrieval methods, and four ODQA benchmarks. |
AIR-Bench: Automated Heterogeneous Information Retrieval Benchmark (2025.acl-long)
Copied to clipboard
| Challenge: | Evaluation benchmarks based on predefined domains and human-labeled data face limitations in addressing evaluation needs for emerging domains. |
| Approach: | They propose an automated information retrieval benchmark based on predefined domains and human-labeled data . AIR-Bench is automated and Heterogeneous with three key features . |
| Outcome: | The proposed benchmarks are based on predefined domains and human-labeled data. |
We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning? (2025.acl-long)
Copied to clipboard
Runqi Qiao, Qiuna Tan, Guanting Dong, MinhuiWu MinhuiWu, Chong Sun, Xiaoshuai Song, Jiapeng Wang, Zhuoma GongQue, Shanglin Lei, YiFan Zhang, Zhe Wei, Miaoxuan Zhang, Runfeng Qiao, Xiao Zong, Yida Xu, Peiqing Yang, Zhimin Bao, Muxi Diao, Chen Li, Honggang Zhang
| Challenge: | Existing benchmarks focus more on end-to-end performance, but neglect the underlying principles of knowledge acquisition and generalization. |
| Approach: | They propose a benchmark specifically designed to explore the problem-solving principles by decomposing 6.5K visual math problems into 10.9K step-level questions for evaluation. |
| Outcome: | The proposed benchmark covers 6.5K visual math problems and 10.9K step-level questions spanning 5 layers of knowledge granularity and 67 hierarchical knowledge concepts. |
Modeling the Evolution of English Noun Compounds with Feature-Rich Diachronic Compositionality Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Empirical research directly addressing these issues is limited to a small number of studies suggesting that compounding is a highly productive process. |
| Approach: | They represent English noun compounds as vectors of time-specific values and implement a set of features to classify them for present-day compositionality and assess the informativeness of the corresponding linguistic patterns. |
| Outcome: | The proposed method captures relevant and complementary information across approaches and shows that low-compositional meanings are reflected by a parallel drop in compositionality and sustained semantic change. |
What’s the Difference? Supporting Users in Identifying the Effects of Prompt and Model Changes Through Token Patterns (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods for prompting for large language models have limitations such as being labor-intensive or lacking insights. |
| Approach: | They propose a new approach that automatically distinguishes between random variations and systematic differences in language model outputs by using token patterns. |
| Outcome: | The proposed method combines both automation and human analysis to provide new insights into established prompt data. |
V-Oracle: Making Progressive Reasoning in Deciphering Oracle Bones for You and Me (2025.acl-long)
Copied to clipboard
Runqi Qiao, Qiuna Tan, Guanting Dong, MinhuiWu MinhuiWu, Jiapeng Wang, YiFan Zhang, Zhuoma GongQue, Chong Sun, Yida Xu, Yadong Xue, Ye Tian, Zhimin Bao, Lan Yang, Chen Li, Honggang Zhang
| Challenge: | Deciphering oracle bone scripts using AI technology is not an overnight task due to the evolution of written language over millennia. |
| Approach: | They propose a framework that utilizes Large Multi-modal Models (LMMs) for interpreting Oracle Bone Script (OBS). |
| Outcome: | The proposed framework provides quantitative analyses and superior deciphering capability. |
Unveiling Cultural Blind Spots: Analyzing the Limitations of mLLMs in Procedural Text Comprehension (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated exceptional capabilities across various natural language processing tasks, including text summarization, multi-modal machine translation, and code generation and understanding. |
| Approach: | They propose a benchmark to evaluate mLLMs’ ability to process and reason over culturally diverse procedural texts in multiple languages. |
| Outcome: | The proposed benchmarks show that mLLMs struggle with culturally contextualized procedural content, especially in low-resource languages, and perform better on multiple-choice tasks presented in conversational formats than on direct questions. |
Improving Language and Modality Transfer in Translation by Character-level Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Current translation systems cover only 5% of the world's languages . expanding to the long-tail of low-resource languages requires data-efficient methods that rely on cross-lingual and cross-modal knowledge transfer. |
| Approach: | They propose a character-based approach to improve adaptability to new languages and modalities by using a teacher-student approach and parallel translation data to obtain a SONAR character-level encoder. |
| Outcome: | The proposed model outperforms subword-based models in speech-to-text translation on the FLEURS benchmark on 33 languages and achieves state-of-the-art generalizability to unseen languages. |
DialUp! Modeling the Language Continuum by Adapting Models to Dialects and Dialects to Models (2025.acl-long)
Copied to clipboard
Niyati Bafna, Emily Chang, Nathaniel Romney Robinson, David R. Mortensen, Kenton Murray, David Yarowsky, Hale Sirin
| Challenge: | Recent advances in MT quality and language coverage have shown that language varieties with low baseline performance are more likely to benefit from these approaches. |
| Approach: | They propose a training-time technique for adapting a pretrained model to dialectal data and an inference-time intervention adapting dialectal datasets to the model expertise. |
| Outcome: | The proposed model shows significant performance gains for several dialects from four language families, and modest gains for two other language families. |
AutoMixAlign: Adaptive Data Mixing for Multi-Task Preference Optimization in LLMs (2025.acl-long)
Copied to clipboard
Nicholas E. Corrado, Julian Katz-Samuels, Adithya M Devraj, Hyokun Yun, Chao Zhang, Yi Xu, Yi Pan, Bing Yin, Trishul Chilimbi
| Challenge: | Existing approaches to align large language models rely on large ablation studies, heuristics, or human intuition to produce models with strong performance across tasks. |
| Approach: | They propose an algorithm that mixes datasets during LLM training to balance performance across multiple tasks. |
| Outcome: | The proposed algorithm outperforms existing methods on multitask alignment setups and achieves convergence rate of O(1/T) in the convex case. |
Modeling Complex Semantics Relation with Contrastively Fine-Tuned Relational Encoders (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for learning relational embeddings fail to capture nuanced representations and rich semantics. |
| Approach: | They propose different relational encoders designed to capture diverse relational aspects and semantic properties of entity pairs. |
| Outcome: | The proposed encoders capture diverse relational aspects and semantic properties of entity pairs. |
Error-driven Data-efficient Large Multimodal Model Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Large Multimodal Models (LMMs) have demonstrated impressive performance across numerous academic benchmarks, but task-specific tuning samples are often not readily available or expensive and time-consuming to obtain. |
| Approach: | They propose an error-driven data-efficient tuning framework that aims to efficiently adapt generic LMMs to newly emerging tasks without extensive task-specific training samples. |
| Outcome: | The proposed model achieves an average performance boost of 7.01% on seven tasks across three training data scales and three different training datascales. |
Planning with Diffusion Models for Target-Oriented Dialogue Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for directing conversations toward specific targets generate dialogue plans in a step-by-step sequential manner and suffer from compounding errors and myopic actions. |
| Approach: | They propose a framework that leverages diffusion models to enable non-sequential dialogue planning. |
| Outcome: | The proposed framework performs non-myopic lookahead exploration and optimizes action strategies over a long horizon through non-sequential dialogue planning. |
Interactive and Expressive Code-Augmented Planning with Large Language Models (2025.acl-long)
Copied to clipboard
Anthony Zhe Liu, Xinhe Wang, Jacob Sansom, Yao Fu, Jongwook Choi, Sungryull Sohn, Jaekyeom Kim, Honglak Lee
| Challenge: | Large Language Models (LLMs) have strong abilities in common-sense reasoning and interactive decision-making, but struggle with complex, long-horizon planning tasks. |
| Approach: | They propose a code-based LLM planning approach that is code-expressive while also dynamically adapting from errors. |
| Outcome: | The proposed approach can be error-prone and insufficient for handling ambiguous or unstructured data. |
Synergistic Weak-Strong Collaboration by Aligning Preferences (2025.acl-long)
Copied to clipboard
Yizhu Jiao, Xuchao Zhang, Zhaoyang Wang, Yubo Ma, Zhun Deng, Rujia Wang, Chetan Bansal, Saravan Rajmohan, Jiawei Han, Huaxiu Yao
| Challenge: | Current Large Language Models excel in general reasoning yet struggle with specialized tasks requiring proprietary or domain-specific knowledge. |
| Approach: | They propose a collaborative framework that pairs a specialized weak model with a general strong model to optimize collaboration. |
| Outcome: | The proposed framework outperforms each model alone by leveraging complementary strengths. |
Understanding Silent Data Corruption in LLM Training (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are a challenging task because of their large size and complexity. |
| Approach: | They propose to isolate and analyze the impact of SDCs on LLM training by using a cloud computing platform to access unhealthy nodes swept out of production by automated fleet management. |
| Outcome: | The proposed model training compares healthy production nodes with unhealthy nodes exhibiting SDCs at three levels: at each submodule computation, at a single optimizer step, and at . training period. |
Align-SLM: Textless Spoken Language Models with Reinforcement Learning from AI Feedback (2025.acl-long)
Copied to clipboard
Guan-Ting Lin, Prashanth Gurunath Shivakumar, Aditya Gourav, Yile Gu, Ankur Gandhe, Hung-yi Lee, Ivan Bulyko
| Challenge: | Textless Spoken Language Models lag behind text-based Large Language Model (LLM) in semantic coherence and relevance. |
| Approach: | They propose a framework that leverages preference optimization inspired by Reinforcement Learning with Human Feedback to enhance the semantic understanding of SLMs. |
| Outcome: | The proposed framework achieves state-of-the-art performance of SLMs for most benchmarks . it leverages preference optimization inspired by Reinforcement Learning with Human Feedback . |
Can LLMs Help Uncover Insights about LLMs? A Large-Scale, Evolving Literature Analysis of Frontier LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent surveys of literature highlight the overwhelming growth of Large Language Models (LLMs). |
| Approach: | They propose a semi-automated literature analysis approach that automates literature analysis using LLMs. |
| Outcome: | The proposed approach reduces paper surveying and data extraction by 93% compared to manual methods. |
BIG5-CHAT: Shaping LLM Personalities Through Training on Human-Grounded Data (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for embedding human personality traits into LLMs are limited by realism and validity issues. |
| Approach: | They propose to use a large-scale dataset to embed human personality traits into LLMs . they use supervised fine-tuning and direct preference optimization to train LLM models . |
| Outcome: | The proposed methods outperform prompting on personality assessments and IPIP-NEO, and show higher conscientiousness, agreeableness, lower extraversion, and lower neuroticism on reasoning tasks. |
Deep Temporal Reasoning in Video Language Models: A Cross-Linguistic Evaluation of Action Duration and Completion through Perfect Times (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that video-language models struggle to mirror human-like temporal reasoning grounded in video . Sequential events are not simply arranged chronologically; rather, one event triggers the next upon reaching its completion. |
| Approach: | They propose a quadrilingual dataset to assess temporal reasoning in video-language models . they pair everyday activity videos with event completion labels and perfectivity distractors . |
| Outcome: | The perfect times dataset examines whether video-language models comprehend temporal dynamics . it combines everyday activity videos with event completion labels and perfectivity distractors . results show that state-of-the-art models struggle to mirror human-like temporal reasoning . |
Amplifying Trans and Nonbinary Voices: A Community-Centred Harm Taxonomy for LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on harms of language technology to transgender and nonbinary people focus on misgendering and stereotyping . |
| Approach: | They propose a taxonomy of harms for large language models and heuristics for evaluation to help identify harmful behavior in LLMs. |
| Outcome: | The proposed model-based approach combines surveys and focus groups with community experts to identify harmful behavior in large language models. |
Enhancing Human Evaluation in Machine Translation with Comparative Judgement (2025.acl-long)
Copied to clipboard
| Challenge: | Human evaluation is crucial for assessing rapidly evolving language models but is influenced by annotator proficiency and task design. |
| Approach: | They evaluate three annotation setups to integrate comparative judgment into human annotation for machine translation. |
| Outcome: | The proposed approach improves inter-annotator agreement and stability of the annotations. |
Infogen: Generating Complex Statistical Infographics from Documents (2025.acl-long)
Copied to clipboard
| Challenge: | Existing efforts to generate simple charts have focused on generating simple infographics from text-heavy documents. |
| Approach: | They propose to generate statistical infographics composed of multiple sub-charts that are contextually accurate, insightful, and visually aligned. |
| Outcome: | The proposed framework outperforms both open-source and closed LLMs in text-to-statistical infographic generation. |
Partial Colexifications Improve Concept Embeddings (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for embedding words from colexification networks are limited to the word level, ignoring lexical relations that would only hold for parts of words in a given language. |
| Approach: | They propose to embed concepts from automatically constructed colexification networks . they use lexical similarity ratings and word association data to evaluate the methods . |
| Outcome: | The proposed methods capture and represent different semantic relationships between concepts. |
Improved Unbiased Watermark for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Unbiased watermarks allow to distinguish between text generated by humans and machines without causing distortion. |
| Approach: | They introduce a family of unbiased, Multi-Channel-based watermarks that partition the language model into segments and promote token probabilities within a selected segment based on a watermark key. |
| Outcome: | The proposed watermarks preserve the original distribution of the language model and offer significant improvements in detectability and robustness over existing unbiased watermark systems. |
MaCP: Minimal yet Mighty Adaptation via Hierarchical Cosine Projection (2025.acl-long)
Copied to clipboard
| Challenge: | MaCP is a new adaptation method for large foundation models that requires minimal parameters and memory for fine-tuning. |
| Approach: | They propose a method that exploits the superior energy compaction and decorrelation properties of cosine projection to improve model efficiency and accuracy. |
| Outcome: | The proposed method improves model efficiency and accuracy across a wide range of single-modality tasks including natural language understanding, natural language generation, text summarization, and multi-modalities such as image classification and video understanding. |
Multi-Attribute Steering of Language Models via Targeted Intervention (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches for steering large language models fail to scale to multi-attribute settings with conflicts, such as enhancing helpfulness while also reducing toxicity. |
| Approach: | They propose a steering framework for selective token-level intervention across multiple attributes that enforcing sparsity and orthogonality among vectors for different attributes. |
| Outcome: | The proposed framework outperforms existing ITI and parameter-efficient fine-tuning approaches across question answering tasks and generative tasks. |
AdaptAgent: Adapting Multimodal Web Agents with Few-Shot Learning from Human Demonstrations (2025.acl-long)
Copied to clipboard
| Challenge: | State-of-the-art multimodal web agents can perform many web tasks by processing user instructions and interacting with graphical user interfaces (GUIs). |
| Approach: | They propose to build multimodal web agents for few-shot adaptability using human demonstrations to improve their generalization and adaptability. |
| Outcome: | The proposed framework enables both proprietary and open-weights multimodal web agents to adapt to new websites and domains using few human demonstrations. |
Can LLMs Identify Critical Limitations within Scientific Research? A Systematic Evaluation on AI Research Papers (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have demonstrated remarkable capabilities across a variety of scientific tasks, such as answering questions about scientific papers, writing scientific papers and retrieving related works. |
| Approach: | They propose a taxonomy of limitation types in scientific research with a focus on AI to evaluate their ability to support early-stage feedback and complement human peer review. |
| Outcome: | The proposed model enhances the ability of LLM systems to generate limitations in research papers, enabling them to provide more concrete and constructive feedback. |
On the Acquisition of Shared Grammatical Representations in Bilingual Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Crosslingual transfer is crucial to contemporary language models’ multilingual capabilities, but how it occurs is not well understood. |
| Approach: | They use structural priming to study grammatical representations in humans by controlling for training data quantity and language exposure. |
| Outcome: | The proposed model is able to learn a language in two languages and has a higher likelihood of learning a prepositional object (PO) dative sentence than a double object (DO) . |
Using Shapley interactions to understand how models use structure (2025.acl-long)
Copied to clipboard
| Challenge: | Language models are intricately structured systems, and attribution measures are important for understanding how they combine features to influence outputs. |
| Approach: | They use Shapley Taylor interaction indices to examine how language and speech models internally relate and structure their inputs. |
| Outcome: | The proposed methods show that language models encode phonetic interactions . they show that the inputs are more entangled for pairs where a consonant influences a vowel or approximant . |
Adversarial Tokenization (2025.acl-long)
Copied to clipboard
| Challenge: | Current LLM pipelines account for only one possible tokenization for a given string . authors: noncanonical tokenizations can evade LLM safety while still generating meaningful responses. |
| Approach: | They show that LLM pipelines account for only one possible tokenization for a given string . they show that tokenizers retain semantic understanding of other tokenizations . authors propose an exploit that can be exploited to evade safety and alignment restrictions . |
| Outcome: | The proposed exploit exploits a previously unknown vulnerability in subword models. |
Classifying Unreliable Narrators with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study identifies unreliable narrators, i.e. those who unintentionally misrepresent information . authors propose using computational methods to identify unredependable narrators . adbrei: readers implicitly question the reliability of the nrator . |
| Approach: | They propose using computational methods to identify unreliable narrators . they use literary theory to define different types of unredependable narrators . |
| Outcome: | The proposed method can identify unreliable narrators on real-world text data. |
ConceptCarve: Dynamic Realization of Evidence (2025.acl-long)
Copied to clipboard
| Challenge: | ConceptCarve is a social media-based evidence retrieval framework that enables users to analyze complex thought patterns that manifest differently across social media communities. |
| Approach: | They propose ConceptCarve, an evidence retrieval framework that utilizes traditional retrieval systems and LLMs to dynamically characterize the search space during retrieval. |
| Outcome: | The proposed framework surpasses traditional retrieval systems in finding evidence within a social media community and produces an interpretable representation of the evidence for that community. |
QQSUM: A Novel Task and Model of Quantitative Query-Focused Summarization for Review-based Product Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing review-based product question answering systems generate only a single answer, ignoring the diversity of viewpoints. |
| Approach: | They propose a task which aims to summarize diverse customer opinions into representative Key Points and quantify their prevalence to effectively answer user queries. |
| Outcome: | The proposed task summarizes diverse customer opinions into representative Key Points and quantifies their prevalence to answer user queries. |
Navigating Rifts in Human-LLM Grounding: Study and Benchmark (2025.acl-long)
Copied to clipboard
| Challenge: | Language models excel at following instructions but struggle with collaborative aspects of conversation that humans naturally employ. |
| Approach: | They analyze logs from WildChat, MultiWOZ, and Bing Chat to examine grounding challenges . they propose a benchmark to determine when LLMs fail to initiate grounding . |
| Outcome: | The proposed model predicts interactions that fail to ground with users . the proposed model is based on human-human interactions with humans . |
Substance over Style: Evaluating Proactive Conversational Coaching Agents (2025.acl-long)
Copied to clipboard
Vidya Srinivas, Xuhai Xu, Xin Liu, Kumar Ayush, Isaac Galatzer-Levy, Shwetak Patel, Daniel McDuff, Tim Althoff
| Challenge: | Recent NLP research has focused on single-turn tasks with well-defined objectives or evaluation criteria. |
| Approach: | They describe five multi-turn coaching agents that exhibit distinct conversational styles and evaluate them through a user study. |
| Outcome: | The authors compare user feedback with third-person evaluations from health experts and an LM to find that stylistic components in absence of core functionality are viewed negatively. |
Open-World Planning via Lifted Regression with LLM-Inferred Affordances for Embodied Agents (2025.acl-long)
Copied to clipboard
Xiaotian Liu, Ali Pesaranghader, Hanze Li, Punyaphat Sukcharoenchaikul, Jaehong Kim, Tanmana Sadhu, Hyejeong Jeon, Scott Sanner
| Challenge: | Existing open-world planning methods rely on closed-world assumption (CWA) symbolic planners face combinatorial explosion of states and actions due to reliance on grounding. |
| Approach: | They propose an open-world planning approach integrating lifted regression with LLM-generated affordances. |
| Outcome: | The proposed approach outperforms state-of-the-art LLM planners and a grounded planner on three benchmarks. |
(RSA)²: A Rhetorical-Strategy-Aware Rational Speech Act Framework for Figurative Language Understanding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing implementations of the Rational Speech Act (RSA) framework do not account for figurative expressions or require modeling the implicit motivations behind using figurativ language in a setting-specific way. |
| Approach: | They propose a framework which models figurative language use by considering a speaker's employed rhetorical strategy and a computational model which incorporates rhetorical strategies to support non-literal interpretation. |
| Outcome: | The proposed framework enables human-compatible interpretations of non-literal utterances without modeling speaker's motivations for being non-lative. |
SYNTHIA: Novel Concept Design with Affordance Composition (2025.acl-long)
Copied to clipboard
Hyeonjeong Ha, Xiaomeng Jin, Jeonghwan Kim, Jiateng Liu, Zhenhailong Wang, Khanh Duy Nguyen, Ansel Blume, Nanyun Peng, Kai-Wei Chang, Heng Ji
| Challenge: | Existing studies on concept design using text-to-image models have enabled rapid ideation of novel visual concepts. |
| Approach: | They propose a framework for generating novel, functionally coherent designs based on desired affordances by decomposing concepts into parts and affordance . they also develop a curriculum learning scheme that fine-tunes T2I models to progressively learn affordance composition while maintaining visual novelty. |
| Outcome: | The proposed framework outperforms state-of-the-art models for novelty and functional coherence in human evaluation. |
Consistent Client Simulation for Motivational Interviewing-based Counseling (2025.acl-long)
Copied to clipboard
Yizhe Yang, Palakorn Achananuparp, Heyan Huang, Jing Jiang, Nicholas Gabriel Lim, Cameron Tan Shi Ern, Phey Ling Kit, Jenny Giam Xiuhui, John Pinto, Ee-Peng Lim
| Challenge: | Existing approaches to simulate human clients in mental health counseling are limited and cost prohibitive. |
| Approach: | They propose a framework that supports consistent client simulation for mental health counseling by tracking the mental state of a simulated client, controlling its state transitions, and generating for each state behaviors consistent with the client’s motivation, beliefs, preferred plan to change, and and receptivity. |
| Outcome: | The proposed framework can simulate human clients for mental health counseling tasks and achieve higher consistency than previous methods. |
AUTALIC: A Dataset for Anti-AUTistic Ableist Language In Context (2025.acl-long)
Copied to clipboard
Naba Rizvi, Harper Strickland, Daniel Gitelman, Alexis Morales Flores, Tristan Cooper, Aekta Kallepalli, Akshat Alurkar, Haaset Owens, Saleha Ahmedi, Isha Khirwadkar, Imani N. S. Munyaka, Nedjma Ousidhoum
| Challenge: | Existing tools for detecting anti-autistic ableist language are lacking in this domain . current tools fail to accurately identify anti- autistic language, despite its subtle nature . |
| Approach: | They present a dataset dedicated to the detection of anti-autistic ableist language in context . they use reddit sentences with surrounding context to identify anti-ableist expressions . |
| Outcome: | AUTALIC is the first dataset dedicated to the detection of anti-autistic ableist language in context . it includes 2,400 autism-related sentences collected from reddit and annotated by trained experts . |
Structural Reasoning Improves Molecular Understanding of LLM (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have shown significant performance, approaching human perception levels. |
| Approach: | They propose an approach that sketches molecular structures for reasoning by explicitly incorporating key structural features into the model. |
| Outcome: | The proposed framework improves molecular understanding through extensive experiments. |
CAMI: A Counselor Agent Supporting Motivational Interviewing through State Inference and Topic Exploration (2025.acl-long)
Copied to clipboard
Yizhe Yang, Palakorn Achananuparp, Heyan Huang, Jing Jiang, Phey Ling Kit, Nicholas Gabriel Lim, Cameron Tan Shi Ern, Ee-Peng Lim
| Challenge: | Motivational Interviewing (MI) is a client-centered counseling technique designed to address ambivalence and facilitate behavior change in clients. |
| Approach: | They propose to use a STAR framework to evoke change talk by using large language models to assess MI skill competency, client’s state inference accuracy, topic exploration proficiency, and overall counseling success. |
| Outcome: | The proposed agent outperforms several state-of-the-art methods and shows more realistic counselor-like behavior. |
Know You First and Be You Better: Modeling Human-Like User Simulators via Implicit Profiles (2025.acl-long)
Copied to clipboard
| Challenge: | Existing user simulators lack authenticity and user-level diversity in interactions with large language models. |
| Approach: | They propose a user simulator with implicit user profiles that infers user profiles from human-machine interactions to simulate personalized and realistic dialogues. |
| Outcome: | The proposed framework outperforms baselines in authenticity and diversity while maintaining comparable consistency. |
Targeted Syntactic Evaluation for Grammatical Error Correction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation datasets based on learner-produced texts are insufficient for evaluating models . Currently, sequence-to-sequence models and sequence tagging models perform well on beginner-level grammar items . |
| Approach: | They propose a new evaluation paradigm that assesses GEC models using minimal pairs of ungrammatical and grammatically paired sentences for each grammar item. |
| Outcome: | The proposed evaluation paradigm assesses models using minimal pairs of ungrammatical and grammatically-spaced sentences for each grammar item. |
VF-Eval: Evaluating Multimodal LLMs for Generating Feedback on AIGC Videos (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal large language models (MLLMs) are used for video quality assessment, image captioning and video analysis. |
| Approach: | They propose a benchmark to evaluate MLLMs on AIGC videos using coherence validation, error awareness, error type detection and reasoning evaluation tasks. |
| Outcome: | The proposed benchmark evaluates 13 frontier MLLMs on AIGC videos. |
Language Model Fine-Tuning on Scaled Survey Data for Predicting Distributions of Public Opinions (2025.acl-long)
Copied to clipboard
| Challenge: | Prior studies have failed to accurately predict distribution of survey responses from human subjects. |
| Approach: | They propose to fine-tune large language models to predict human response distributions by leveraging unique structural characteristics of survey data. |
| Outcome: | The proposed model can capture group-specific variability in public opinions, generalizing to unseen subpopulations, survey waves and question topics, and different survey families. |
TESS 2: A Large-Scale Generalist Diffusion Language Model (2025.acl-long)
Copied to clipboard
| Challenge: | Existing instruction-following diffusion models are predominantly trained using an autoregressive paradigm. |
| Approach: | They propose a general instruction-following diffusion language model that outperforms contemporary instruction-tuned diffusion models and matches and sometimes exceeds strong autoregressive (AR) models. |
| Outcome: | The proposed model outperforms and sometimes exceeds existing autoregressive (AR) models on a number of tasks. |
KatFishNet: Detecting LLM-Generated Korean Text through Linguistic Feature Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Detecting LLM-generated text is crucial for academic integrity, preventing plagiarism, protecting copyrights, ethical research practices. |
| Approach: | They propose a method specifically designed for Korean language to detect LLM-generated text . they examine spacing patterns, part-of-speech diversity, and comma usage . |
| Outcome: | The proposed method achieves an average of 19.78% higher AUC-ROC compared to the best-performing detection method. |
Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQL (2025.acl-long)
Copied to clipboard
| Challenge: | Direct Preference Optimization (DPO) is effective in complex reasoning tasks like math word problems and code generation, but Text-to-SQL datasets often include only final answers (gold SQL queries) without detailed CoT solutions. |
| Approach: | They found that Direct Preference Optimization (DPO) is crucial for unlocking DPO's potential by augmenting Text-to-SQL datasets with synthetic CoT solutions. |
| Outcome: | The proposed method achieves consistent and significant performance improvements on Text-to-SQL datasets. |
On Generalization across Measurement Systems: LLMs Entail More Test-Time Compute for Underrepresented Cultures (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) should be able to provide accurate information irrespective of the measurement system at hand . |
| Approach: | They use newly compiled datasets to test if this is true for seven open-source LLMs. |
| Outcome: | The proposed model can provide accurate information regardless of the measurement system at hand. |
CORDIAL: Can Multimodal Large Language Models Effectively Understand Coherence Relationships? (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks focus on assessing factual and logical correctness in downstream tasks with limited emphasis on evaluating MLLMs’ ability to interpret pragmatic cues and intermodal relationships. |
| Approach: | They propose to use Coherence Relations to assess MLLMs' ability to perform multimodal discourse analysis using different prompting strategies. |
| Outcome: | The proposed model fails to match the performance of simple classifier-based benchmarks on 10+ MLLMs using different prompting strategies. |
Veracity Bias and Beyond: Uncovering LLMs’ Hidden Beliefs in Problem-Solving Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have been aligned to avoid harmful biases and stereotypes, but recent studies have revealed the superficial nature of this alignment. |
| Approach: | They propose to use large language models to avoid harmful biases and stereotypes by assigning personas to LLMs to observe decision discrepancies in social scenarios or asking them to associate specific attributes with social targets. |
| Outcome: | The proposed models attribute fewer correct solutions and more incorrect ones to African-American groups in math and coding, while Asian authorships are least preferred in writing evaluation. |
Optimal Transport-Based Token Weighting scheme for Enhanced Preference Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for direct preference optimization assign equal importance to all tokens while humans focus on more meaningful parts. |
| Approach: | They propose to use a transport-based token weighting scheme to enhance direct preference optimization by emphasizing meaningful token pairs and de-emphasizing less relevant ones to yield a more contrastive reward difference estimate. |
| Outcome: | Extensive experiments have validated the proposed method in improving instruction-following ability across various settings. |
LLM Meets Scene Graph: Can Large Language Models Understand and Generate Scene Graphs? A Benchmark and Empirical Study (2025.acl-long)
Copied to clipboard
Dongil Yang, Minjin Kim, Sunghwan Kim, Beong-woo Kwak, Minjun Park, Jinseok Hong, Woontack Woo, Jinyoung Yeo
| Challenge: | Large language models (LLMs) have demonstrated impressive progress in various text-based tasks, such as question-answering and content generation. |
| Approach: | They propose a benchmark to evaluate Large Language Models’ ability to understand scene graphs and generate them from textual narratives. |
| Outcome: | The proposed model performs well on scene graph understanding but struggles with scene graph generation, particularly for complex narratives. |
Beyond Frameworks: Unpacking Collaboration Strategies in Multi-Agent Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Existing frameworks prioritize structural architectures and role assignments but neglect granular mechanics of agent collaboration. |
| Approach: | They propose to use centralized governance, instructor-led participation, ordered interaction patterns to optimize task accuracy and computational efficiency. |
| Outcome: | The proposed model improves task accuracy and computational efficiency under two context-dependent scenarios. |
The Invisible Hand: Unveiling Provider Bias in Large Language Models for Code Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have emerged as the new recommendation engines, surpassing traditional methods in both capability and scope, particularly in code generation. |
| Approach: | They propose to use a dataset to investigate a new type of bias in Large Language Models for code generation, provider bias, to determine whether the model favors specific providers. |
| Outcome: | The proposed model favors services from Google and Amazon, but without explicit directives, and can modify input code to incorporate their preferred providers without user requests. |
K/DA: Automated Data Generation Pipeline for Detoxifying Implicitly Offensive Language in Korean (2025.acl-long)
Copied to clipboard
| Challenge: | Language detoxification involves removing toxicity from offensive language. |
| Approach: | They propose an automated pipeline to generate offensive language with implicit offensiveness and trend-aligned slang. |
| Outcome: | The proposed dataset exhibits high pair consistency and greater implicit offensiveness compared to existing Korean datasets and demonstrates applicability to other languages. |
THOR-MoE: Hierarchical Task-Guided and Context-Responsive Routing for Neural Machine Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing sparse Mixture-of-Experts (MoE) solutions may lead to sub-optimal performance . thor-moe uses domain/linguistics-specific knowledge, but lacks context-responsive routing policies . |
| Approach: | They propose a sparse Mixture-of-Experts (MoE) solution which uses task knowledge of NMT into MoE and provides hierarchical task-guided and context-responsive routing policies. |
| Outcome: | thor-MoE can achieve an average improvement of 0.75 BLEU with less than 22% activated parameters on multi-domain translation tasks. |
Neuron Empirical Gradient: Discovering and Quantifying Neurons’ Global Linear Controllability (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have shown that feed-forward neurons in pre-trained language models (PLMs) can encode factual knowledge, but current methods are costly and lack the link between activations and outputs. |
| Approach: | They propose to compute a global linear relationship between neuron activations and outputs using a knowledge probing dataset. |
| Outcome: | The proposed method exploits the neural empirical gradient to capture changes in neuron activations and model outputs. |
Can Third Parties Read Our Emotions? (2025.acl-long)
Copied to clipboard
Jiayi Li, Yingfan Zhou, Pranav Narayanan Venkit, Halima Binte Islam, Sneha Arya, Shomir Wilson, Sarah Rajtmajer
| Challenge: | Existing approaches to infer author’s private states from written text have relied heavily on datasets annotated by third-party annotators. |
| Approach: | They propose a framework for evaluating the limitations of third-party annotations and call for refined annotation practices to accurately represent and model authors’ private states. |
| Outcome: | The proposed methods outperform human annotators on emotion recognition tasks. |
OZSpeech: One-step Zero-shot Speech Synthesis with Learned-Prior-Conditioned Flow Matching (2025.acl-long)
Copied to clipboard
| Challenge: | Text-to-speech systems have seen significant advances in recent years, driven by improvements in deep learning and neural network architectures. |
| Approach: | They propose a method to explore optimal transport conditional flow matching with one-step sampling and a learned prior as the condition, effectively disregarding preceding states and reducing the number of sampling steps. |
| Outcome: | The proposed method achieves promising performance over existing methods in content accuracy, naturalness, prosody generation, and speaker style preservation. |
World Modeling Makes a Better Planner: Dual Preference Optimization for Embodied Task Planning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches focus on action selection or use pre-trained models as world models to enhance planning capabilities. |
| Approach: | They propose a new learning framework that optimizes state prediction and action selection through preference learning. |
| Outcome: | The proposed method outperforms existing methods and GPT-4o on VoTa-Bench and Qwen2-VL (7B), LLaVA-1.6 (7B) and LLama-3.2 (11B). |
JailbreakRadar: Comprehensive Assessment of Jailbreak Attacks Against LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have been used to mitigate misuse and to align with human values. |
| Approach: | They propose to use large-scale evaluations of various jailbreak attacks to identify key patterns and test them under eight advanced defenses. |
| Outcome: | The proposed attacks achieve high success rates but are easy to mitigate by defenses. |
CogniBench: A Legal-inspired Framework and Dataset for Assessing Cognitive Faithfulness of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks focus on “factual statements” that rephrase source materials, but ignore “cognitive statements” . evaluating and detecting "faithfulness hallucinations" remains challenging . |
| Approach: | They propose a framework to assess faithfulness of cognitive statements and introduce a dataset to scale easily across models. |
| Outcome: | The proposed framework assesses faithfulness of cognitive statements and scales easily across models. |
Neural Incompatibility: The Unbridgeable Gap of Cross-Scale Parametric Knowledge Transfer in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have a transparent brain with accessible parameters that encode extensive knowledge, which can be analyzed, located and transferred. |
| Approach: | They propose a new paradigm that aligns parametric spaces of LLMs using several training steps without following training. |
| Outcome: | The proposed model aligns parametric spaces across scales using only training steps without following training. |
Enhancing Mathematical Reasoning in LLMs by Stepwise Correction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Best-of-N decoding methods often lead to incorrect solutions . a novel method is proposed to help large language models identify and revise incorrect steps in their generated reasoning paths. |
| Approach: | They propose a method that helps large language models identify and revise incorrect steps in their generated reasoning paths. |
| Outcome: | The proposed method outperforms the state-of-the-art Best-ofN decoding method by +2.4 and reduces token consumption by 77.8%. |
PsyDial: A Large-scale Long-term Conversational Dataset for Mental Health Support (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models for mental health counseling use a privacy-preserving data reconstruction method to reconstruct client-counselor dialogues without removing personally identifiable information due to privacy concerns. |
| Approach: | They propose a privacy-preserving data reconstruction method that reconstructs real-world client-counselor dialogues while mitigating privacy concerns. |
| Outcome: | The proposed method reduces privacy risks while maintaining dialogue diversity and conversational exchange while maintaining conversational diversity. |
Enhancing Goal-oriented Proactive Dialogue Systems via Consistency Reflection and Correction (2025.acl-long)
Copied to clipboard
| Challenge: | Unlike traditional dialogue systems, goal-oriented proactive dialogue systems focus on achieving specific objectives by actively guiding and anticipating user needs. |
| Approach: | They propose a model-agnostic two-stage Consistency Reflection and Correction framework that allows the model to reflect on discrepancies between generated responses and dialogue contexts and suggest possible corrections. |
| Outcome: | The proposed framework significantly improves the consistency between generated responses and dialogue contexts on three datasets. |
Exclusion of Thought: Mitigating Cognitive Load in Large Language Models for Enhanced Reasoning in Multiple-Choice Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | Multiple-choice questions (MCQs) are widely used and vital assessment format for evaluating large language models (LLMs). |
| Approach: | They propose a reasoning prompt strategy that redirects the model's attention away from erroneous options and eliminates incorrect options. |
| Outcome: | The proposed reasoning prompt reduces cognitive load by steering the model’s attention away from erroneous options, enabling the model to focus more effectively on reasonable answers. |
Registering Source Tokens to Target Language Spaces in Multilingual Neural Machine Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Multilingual neural machine translation (MNMT) aims for arbitrary translations across multiple languages. |
| Approach: | They propose a method that inserts a set of tokens specifying the target language into the input sequence between the source and target tokens. |
| Outcome: | The proposed method outperforms existing models on a large-scale benchmark. |
VisuoThink: Empowering LVLM Reasoning with Multimodal Tree Search (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to large vision-language models fail to capture interleaved nature of human visual-verbal reasoning processes. |
| Approach: | They propose a framework that integrates visuospatial and linguistic domains to facilitate multimodal slow thinking by enabling progressive visual-textual reasoning. |
| Outcome: | Experiments show that VisuoThink significantly improves reasoning capabilities even without fine-tuning. |
Automated CAD Modeling Sequence Generation from Text Descriptions via Transformer-Based Large Language Models (2025.acl-long)
Copied to clipboard
JianXing Liao, Junyan Xu, Yatao Sun, Maowen Tang, Sicheng He, Jingxian Liao, Shui Yu, Yun Li, Xiaohong Guan
| Challenge: | Experimental results demonstrate that the proposed approach outperforms traditional methods in both accuracy and efficiency. |
| Approach: | They propose a language-guided framework that integrates large language models with computer-automated design to address these challenges. |
| Outcome: | The proposed framework outperforms traditional methods in accuracy and efficiency, providing a powerful tool for automating industrial workflows and generating complex CAD models from textual prompts. |
LED-Merging: Mitigating Safety-Utility Conflicts in Model Merging with Location-Election-Disjoint (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning large language models for specialized tasks are costly and time-consuming. |
| Approach: | They propose a framework that locates task-specific neurons via gradient-based attribution and dynamically Elects critical neurons through multi-model importance fusion. |
| Outcome: | The proposed framework reduces harmful response rates while preserving 95% of utility performance. |
Dolphin: Moving Towards Closed-loop Auto-research through Thinking, Practice, and Feedback (2025.acl-long)
Copied to clipboard
Jiakang Yuan, Xiangchao Yan, Bo Zhang, Tao Chen, Botian Shi, Wanli Ouyang, Yu Qiao, Lei Bai, Bowen Zhou
| Challenge: | Recent studies show that AI-assisted research methods can improve research efficiency . a closed-loop framework is used to enhance the automation level of scientific research . |
| Approach: | They propose a closed-loop LLM-driven framework to enhance the automation level of scientific research. |
| Outcome: | The proposed framework improves the efficiency of scientific research by improving data analysis, accelerating computation, and fostering novel idea generation. |
PerSphere: A Comprehensive Framework for Multi-Faceted Perspective Retrieval and Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that the main challenge lies in long context and perspective extraction. |
| Approach: | They propose a benchmark to facilitate multi-faceted perspective retrieval and summarization . they propose measurable metrics to evaluate the comprehensiveness of the retrieval pipeline . |
| Outcome: | The proposed system breaks free from information silos by combining two opposing claims . it can be used to extract multiple perspectives and improve performance on the platform . |
Prompt-Guided Internal States for Hallucination Detection of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) generate incorrect or logically incorrect responses, which is known as LLM hallucinations. |
| Approach: | They propose a framework for supervised hallucination detection using in-domain data by prompting changes to the structure related to text truthfulness in LLMs’ internal states. |
| Outcome: | The proposed framework enhances the cross-domain generalization of existing hallucination detection methods. |
Typology-Guided Adaptation in Multilingual Models (2025.acl-long)
Copied to clipboard
| Challenge: | Multilingual models often treat language diversity as a problem of data imbalance, overlooking structural variation. |
| Approach: | They propose a typologically grounded metric that quantifies how strongly a language relies on morphology for noun classification. |
| Outcome: | The proposed model outperforms baseline models on 10 Bantu languages . it improves Swahili accuracy by 14 points while maintaining performance on morphology-rich languages like Zulu . |
Don’t Erase, Inform! Detecting and Contextualizing Harmful Language in Cultural Heritage Collections (2025.acl-long)
Copied to clipboard
| Challenge: | Cultural Heritage metadata can contain outdated or offensive terms that reflect historical cultural and societal norms. |
| Approach: | They propose an AI-powered tool that detects offensive terms in CH metadata . the tool has processed over 7.9 million records and provides contextual insights . |
| Outcome: | The proposed tool has processed over 7.9 million records and provides contextual insights . it pairs biased language with contextual information and suggestions for appropriate usage . |
ECLM: Entity Level Language Model for Spoken Language Understanding with Chain of Intent (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated impressive capabilities in language generation and general task performance, but their application to spoken language understanding remains challenging. |
| Approach: | They propose an Entity-level Language Model framework which reformulates slot-filling as an entity recognition task and introduces a new concept, Chain of Intent, to enable step-by-step multi-intent recognition. |
| Outcome: | The proposed framework outperforms strong baselines such as Uni-MIS and achieves gains of 3.7% and 3.1% on MixATIS and MixSNIPS. |
FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing faithful RAG approaches enforce strict context adherence, but they forcibly suppress the model’s parametric knowledge, which undermines the model's internal knowledge structure and increases the risk of misinterpreting the context. |
| Approach: | They propose a framework that resolves knowledge conflicts by explicitly modeling discrepancies between the model’s parametric knowledge and retrieved context. |
| Outcome: | The proposed framework outperforms state-of-the-art methods in knowledge conflict cases and identifies conflicting knowledge at the fact level and designs a self-thinking process. |
Knowledge Image Matters: Improving Knowledge-Based Visual Reasoning with Multi-Image Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Knowledge-based visual reasoning (KB-VR) is a challenging task, as it requires machines not only to understand concepts and relationships of visual scenes, but also to associate them with external world knowledge to perform chain of reasoning on open-world questions. |
| Approach: | They propose a visual knowledge card (VKC) that integrates internal visual knowledge and external world knowledge produced by a knowledge generator into an image. |
| Outcome: | The proposed model achieves new state-of-the-art results compared to previous top-performing models on three popular KB-VR benchmarks. |
Evaluating Personalized Tool-Augmented LLMs from the Perspectives of Personalization and Proactivity (2025.acl-long)
Copied to clipboard
| Challenge: | Personalized tool utilization is essential for aligning large language models (LLMs) with user preference in interaction scenarios with various tools. |
| Approach: | They propose a key-point-based LLM evaluation method that mitigates biases by manually annotating key points for each test case and providing them to LLM as the reference. |
| Outcome: | The proposed method mitigates biases in the LLM-as-a-judge system by manually annotating key points for each test case and providing them to LLM as the reference. |
GUICourse: From General Vision Language Model to Versatile GUI Agent (2025.acl-long)
Copied to clipboard
Wentong Chen, Junbo Cui, Jinyi Hu, Yujia Qin, Junjie Fang, Yue Zhao, Chongyi Wang, Jun Liu, Guirong Chen, Yupeng Huo, Yuan Yao, Yankai Lin, Zhiyuan Liu, Maosong Sun
| Challenge: | Graphical User Interfaces (GUIs) are a pivotal medium for human-computer interaction. |
| Approach: | They propose a series of datasets for training visual-based GUI agents using general VLMs. |
| Outcome: | The proposed GUICourse datasets show that even a small-sized GUI agent performs better on GUI tasks. |
Evaluating Visual and Cultural Interpretation: The K-Viscuit Benchmark with Human-VLM Collaboration (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to creating inclusive vision-language models rely on human annotators, making it labor-intensive and creating cognitive burdens. |
| Approach: | They propose a semi-automated framework for constructing cultural VLM benchmarks . they use an annotated sample of Korean culture to generate questions . |
| Outcome: | The proposed framework is based on a Korean culture dataset and shows that open-source models lag behind proprietary ones in understanding Korean culture. |
Maximizing the Effectiveness of Larger BERT Models for Compression (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for capturing large BERT models as teachers do not fully exploit the potential advantages of larger teachers. |
| Approach: | They propose a method that leverages a pretrained teacher model to guide the training of a lightweight student model to enhance knowledge transfer. |
| Outcome: | The proposed method enhances knowledge transfer by leveraging a pretrained teacher model to guide the training of a lightweight student model. |
Can LLMs Reason About Program Semantics? A Comprehensive Evaluation of LLMs on Formal Specification Inference (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly being used to automate programming tasks. |
| Approach: | They propose a benchmark to evaluate LLMs' reasoning abilities on program semantics. |
| Outcome: | The proposed benchmark shows that LLMs perform well with simple control flows but struggle with more complex structures, especially loops, even with advanced prompting. |
HACo-Det: A Study Towards Fine-Grained Machine-Generated Text Detection under Human-AI Coauthoring (2025.acl-long)
Copied to clipboard
| Challenge: | Existing literature focuses on binary, document-level detection, neglecting texts composed jointly by human and LLM contributions. |
| Approach: | They propose to use a dataset to generate human-AI coauthored texts via an automatic pipeline with word-level attribution labels. |
| Outcome: | The proposed method can detect human-AI coauthored texts with a numeric AI ratio. |
IndicSynth: A Large-Scale Multilingual Synthetic Speech Dataset for Low-Resource Indian Languages (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in synthetic speech generation technology have enabled the generation of high-quality synthetic (fake) speech that emulates human voices. |
| Approach: | They propose a dataset that contains 4,000 hours of synthetic speech from 989 target speakers for 12 low-resourced Indian languages. |
| Outcome: | The proposed dataset contains 4,000 hours of synthetic speech from 989 target speakers, including 456 females and 533 males for 12 low-resourced Indian languages. |
Reinforced IR: A Self-Boosting Framework For Domain-Adapted Information Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing retrieval methods struggle with highly specialized situations that require extensive domain expertise. |
| Approach: | They propose a method that integrates additional information from an LLM-based generator to enhance query performance and train the retriever to better discriminate the relevant documents identified by the generator. |
| Outcome: | The proposed method outperforms existing domain adaptation methods by a large margin and leads to substantial improvements in retrieval quality across a wide range of application scenarios. |
CoIR: A Comprehensive Benchmark for Code Information Retrieval Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods and benchmarks for information retrieval are inadequately representing the diversity of code in various domains and tasks. |
| Approach: | They propose a benchmark specifically designed to assess code retrieval capabilities. |
| Outcome: | The proposed benchmark aims to invigorate research in the code retrieval domain . it shares the same data schema as other popular benchmarks like MTEB and BEIR . |
Enhancing Multimodal Retrieval via Complementary Information Extraction and Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on capturing information in multimodal data that is similar to their paired texts, but often ignores the complementary information contained in multimodule data. |
| Approach: | They propose a multimodal retrieval approach that employs Complementary Information Extraction and Alignment to capture complementary information in multimodal data. |
| Outcome: | The proposed approach achieves significant improvements over divide-and-conquer models and universal dense retrieval models. |
JoPA: Explaining Large Language Model’s Generation via Joint Prompt Attribution (2025.acl-long)
Copied to clipboard
| Challenge: | Existing attempts to explain the entire language generation often treat input prompt texts independently, ignoring their combinatorial effects on the follow-up generation. |
| Approach: | They propose a framework for explaining how a few prompt texts collaboratively influences the LLM's complete generation. |
| Outcome: | The proposed explanations demonstrate faithfulness and efficiency of the proposed framework. |
Proxy-Driven Robust Multimodal Sentiment Analysis with Incomplete Data (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on optimizing model structures to handle uncertain missingness, but models still face challenges when dealing with uncertain missing data. |
| Approach: | They propose a data-centric robust multimodal sentiment analysis method, Proxy-Driven Robust Multimodal Fusion, which maps unimodal data to the latent space of Gaussian distributions to capture core features and structure. |
| Outcome: | The proposed method outperforms existing models in noise resistance and achieves state-of-the-art performance on multiple benchmark datasets. |
Not All Terms Matter: Recall-Oriented Adaptive Learning for PLM-aided Query Expansion in Open-Domain Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Open-domain question answering (ODQA) systems typically adopt a retriever-reader architecture, where the retriever finds relevant documents, and the reader extracts or synthesizes answers. |
| Approach: | They propose a method that iteratively adjusts the importance weights of QE terms based on their relevance, refining term distinction and enhancing the separation of relevant terms. |
| Outcome: | The proposed method improves retrieval accuracy and overall performance on four ODQA datasets and five QE methods. |
A Mutual Information Perspective on Knowledge Graph Embedding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge graph embedding techniques suffer from high intra-group similarity, loss of semantic information, and insufficient inference capability, particularly in complex relation patterns such as 1-N and N-1 relations. |
| Approach: | They propose a knowledge graph embedding framework that leverages mutual information maximization to improve the semantic representation of entities and relations. |
| Outcome: | Extensive experiments on benchmark datasets demonstrate the effectiveness of the proposed method, with consistent performance improvements across various baseline models. |
Aligned but Blind: Alignment Increases Implicit Bias by Reducing Awareness of Race (2025.acl-long)
Copied to clipboard
| Challenge: | et al., 2012) show value-aligned language models exhibit stereotypes in word association tasks . ignoring racial nuances can perpetuate subtle biases in LMs . |
| Approach: | They propose a bias mitigation strategy that incentivizes representation of racial concepts in early model layers. |
| Outcome: | The proposed approach incentivizes representation of racial concepts in early model layers . it reduces implicit bias by reducing the number of ambiguous inputs, the authors show . |
IOPO: Empowering LLMs with Complex Instruction Following via Input-Output Preference Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing algorithms to improve the ability of LLMs to follow complex instructions are lacking. |
| Approach: | They propose a benchmark to improve the ability to follow complex instructions by using a IOPO alignment method to take input and output preference into consideration. |
| Outcome: | The proposed algorithm shows 8.15%, 2.18% improvements on in-domain data and 5.91%, 2.83% on out-of-domain datasets compared to SFT and DPO respectively. |
ProMALex: Progressive Modular Adapters for Multi-Jurisdictional Legal Language Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to training language models for each jurisdiction fail to leverage common legal principles beneficial for low-resource settings or risk negative interference from conflicting jurisdictional interpretations. |
| Approach: | They propose a parameter-efficient framework that derives hierarchical relationships across jurisdictions and progressively inserts adapter modules across model layers based on jurisdictional similarity. |
| Outcome: | The proposed framework outperforms fully shared and jurisdiction-specific models on two legal language modeling benchmarks. |
Flipping Knowledge Distillation: Leveraging Small Models’ Expertise to Enhance LLMs in Text Matching (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated remarkable capabilities in acquiring diverse knowledge, making them highly effective across a wide range of tasks. |
| Approach: | They propose a flipped knowledge distillation paradigm where LLM learns from SLM . they propose to reinterpret LLMs as encoder-decoder models using LoRA . |
| Outcome: | The proposed model has been deployed in an online application environment and validated on financial and healthcare benchmarks and real-world applications. |
Disentangling Language and Culture for Evaluating Multilingual Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Extensive evaluations of large language models (LLMs) are conducted on a wide range of models, revealing a notable cultural-linguistic synergy phenomenon, where models exhibit better performance when questions are culturally aligned with the language. |
| Approach: | They propose a Dual Evaluation Framework to comprehensively assess the multilingual capabilities of large language models by decomposing evaluation along dimensions of linguistic medium and cultural context. |
| Outcome: | The proposed framework allows for a nuanced analysis of LLMs’ ability to process questions within both native and cross-cultural contexts cross-lingually. |
Detecting Sockpuppetry on Wikipedia Using Meta-Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to model author-specific sockpuppet detection on Wikipedia are limited in data-scarce settings. |
| Approach: | They propose to use meta-learning to improve model adaptation to a new sockpuppet-group by training models across multiple tasks. |
| Outcome: | The proposed technique improves performance in data-scarce settings by training models across multiple tasks. |
Diversity-oriented Data Augmentation with Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing data augmentation methods focus on increasing sample numbers while neglecting sample distribution diversity, which can lead to model overfitting. |
| Approach: | They propose a data augmentation framework that focuses on sample distribution diversity and trains a large language model as a diverse paraphraser. |
| Outcome: | The proposed framework achieves an average performance gain of 10.52% surpassing the runner-up baseline with more than three percentage points. |
CoreEval: Automatically Building Contamination-Resilient Datasets with Real-World Knowledge toward Reliable LLM Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Publicly available datasets can be used to evaluate performance of large language models . however, contamination of test data can artificially inflate model performance . |
| Approach: | They propose a Contamination-resilient Evaluation strategy that updates data with real-world knowledge. |
| Outcome: | The proposed evaluation strategy can be used to update datasets with real-world knowledge. |
RiOT: Efficient Prompt Refinement with Residual Optimization Tree (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for automatic prompt optimization face two challenges: lack of diversity and semantic drift. |
| Approach: | They propose a framework for automatic prompt optimization that iteratively refines prompts through text gradients and selects the best prompt using perplexity. |
| Outcome: | The proposed framework outperforms existing prompt optimization methods and manual prompting on commonsense, mathematical, logical, temporal, and semantic reasoning benchmarks. |
Caution for the Environment: Multimodal LLM Agents are Susceptible to Environmental Distractions (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that multimodal GUI agents are susceptible to environmental distractions. |
| Approach: | They propose a scenario where both user and agent are benign and environment is not malicious . they implement an adversarial environment injection and analyze the approach to improve faithfulness . |
| Outcome: | The proposed approach improves faithfulness of multimodal large language model agents in a graphical user interface environment. |
Automatic Evaluation for Text-to-image Generation: Task-decomposed Framework, Distilled Training, and Meta-evaluation Benchmark (2025.acl-long)
Copied to clipboard
| Challenge: | Existing MLLMs rely on commercial models such as GPT-4o for evaluations, but they are not universally accessible. |
| Approach: | They propose a task decomposition evaluation framework based on GPT-4o to automatically construct a specialized training dataset to break down the multifaceted evaluation process into simpler sub-tasks. |
| Outcome: | The proposed framework outperforms the current state-of-the-art GPT-4o evaluation framework with over 4.6% improvement in Spearman and Kendall correlations with human judgments. |
Mitigating Lost-in-Retrieval Problems in Retrieval Augmented Multi-Hop Question Answering (2025.acl-long)
Copied to clipboard
| Challenge: | Empirical results show that ChainRAG consistently outperforms baselines in both effectiveness and efficiency. |
| Approach: | They propose a method which sequentially handles each sub-question by completing missing key entities and retrieving relevant sentences from a sentence graph for answer generation. |
| Outcome: | The proposed method outperforms baselines on three multi-hop QA datasets. |
TableLoRA: Low-rank Adaptation on Table Structure Understanding for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Tabular data are crucial in many fields and their understanding by large language models (LLMs) under high parameter efficiency paradigm is important. |
| Approach: | They propose a module that uses 2D LoRA to encode low-rank information on cell positions to improve table serialization and representation of two-dimensional structured information within a one-dimensional sequence. |
| Outcome: | Experiments on four tabular-related datasets show that TableLoRA outperforms vanilla LoRA and surpasses table encoding methods tested in control. |
Condor: Enhance LLM Alignment with Knowledge-Driven Data Synthesis and Refinement (2025.acl-long)
Copied to clipboard
Maosongcao Maosongcao, Taolin Zhang, Mo Li, Chuyu Zhang, Yunxin Liu, Conghui He, Haodong Duan, Songyang Zhang, Kai Chen
| Challenge: | Existing high-quality human-annotated SFT data is a bottleneck for Large Language Models (LLMs). |
| Approach: | They propose a two-stage synthetic data generation framework that incorporates World Knowledge Trees and Self-Reflection Refinement to produce high-quality SFT data at scale. |
| Outcome: | The proposed model fine-tuned on 20K condor-generated samples achieves superior performance compared to instruct model trained with RLHF. |
CulFiT: A Fine-grained Cultural-aware LLM Training Paradigm via Multilingual Critique Data Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models exhibit a specific cultural bias, neglecting values and differences of low-resource regions. |
| Approach: | They propose a culturally-aware training paradigm that leverages multilingual data and fine-grained reward modeling to enhance cultural sensitivity and inclusivity. |
| Outcome: | The proposed model achieves state-of-the-art in cultural alignment and general reasoning. |
Decoding Knowledge Attribution in Mixture-of-Experts: A Framework of Basic-Refinement Collaboration and Efficiency Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing attribution methods for dense models fail to capture dynamic routing-expert interactions in sparse MoE architectures. |
| Approach: | They propose to analyze sparse MoE architectures against dense models to capture dynamic routing-expert interactions. |
| Outcome: | The proposed algorithm shows that sparse models achieve higher efficiency per layer . it also shows that deep Qwen-MoE mitigates expert failures while minimizing complexity . |
ChartLens: Fine-grained Visual Attribution in Charts (2025.acl-long)
Copied to clipboard
| Challenge: | MLLMs suffer from hallucinations, where generated text fails to align with visual inputs. |
| Approach: | They propose a chart attribution algorithm that uses segmentation-based techniques to identify chart objects and employs set-of-marks prompting with MLLMs for fine-grained visual attribution. |
| Outcome: | The proposed algorithm improves fine-grained attributions by 26-66% . |
LESA: Learnable LLM Layer Scaling-Up (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for depth scaling-up rely on empirical heuristic rules for layer duplication, resulting in poor initialization and slower convergence during continual pre-training. |
| Approach: | They propose a method for learning latent parameters between layers by concatenating parameters from each layer and applying Singular Value Decomposition. |
| Outcome: | Experiments show that LESA outperforms baseline models with less than half the cost of existing methods. |
MMRC: A Large-Scale Benchmark for Understanding Multimodal Large Language Model in Real-World Conversation (2025.acl-long)
Copied to clipboard
Haochen Xue, Feilong Tang, Ming Hu, Yexin Liu, Qidong Huang, Yulong Li, Chengzhi Liu, Zhongxing Xu, Chong Zhang, Chun-Mei Feng, Yutong Xie, Imran Razzak, Zongyuan Ge, Jionglong Su, Junjun He, Yu Qiao
| Challenge: | Existing multimodal large language models lack the ability to memorize, recall, and reason in sustained interactions. |
| Approach: | They propose a multimodal real-world conversation benchmark for evaluating open-ended abilities of multimodal large language models. |
| Outcome: | The proposed benchmarks show that the models perform better in open-ended conversations. |
Towards the Law of Capacity Gap in Distilling Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Language model (LM) distillation aims at distilling knowledge in a large teacher LM to a small student one. |
| Approach: | They propose to use the law of capacity gap to distill knowledge from a large teacher to a small student model. |
| Outcome: | The proposed model outperforms other language models on a larger scale by using the law of capacity gap inducted from a preliminary study on small-scale (3B) LMs. |
WhiSPA: Semantically and Psychologically Aligned Whisper with Self-Supervised Contrastive and Student-Teacher Learning (2025.acl-long)
Copied to clipboard
Rajath Rao, Adithya V Ganesan, Oscar Kjell, Jonah Luby, Akshay Raghavan, Scott M. Feltman, Whitney Ringwald, Ryan L. Boyd, Benjamin J. Luft, Camilo J. Ruggero, Neville Ryant, Roman Kotov, H. Schwartz
| Challenge: | Current speech encoding pipelines rely on an additional text-based LM to get robust representations of human communication, even though speech-to-text models often have a LM within. |
| Approach: | They propose to align Whisper's latent space with semantic representations from a text autoencoder and lexically derived embeddings of basic psychological dimensions: emotion and personality. |
| Outcome: | The proposed approach surpasses current speech encoders over self-supervised affective tasks and downstream psychological tasks, achieving an error reduction of 73.4% and 83.8%, respectively. |
Keys to Robust Edits: From Theoretical Insights to Practical Advances (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for modifying parametric memory are prone to inaccuracies due to conflicting or outdated information. |
| Approach: | They propose a plug-and-play module that disentangles editing keys from native model representations and dynamically adjusts keys via contrastive learning to achieve robustness-specificity balance. |
| Outcome: | The proposed method improves over robustness tests by up to 66.4% while maintaining the success rate unaffected. |
Boosting LLM’s Molecular Structure Elucidation with Knowledge Enhanced Tree Search Reasoning (2025.acl-long)
Copied to clipboard
Xiang Zhuang, Bin Wu, Jiyu Cui, Kehua Feng, Xiaotong Li, Huabin Xing, Keyan Ding, Qiang Zhang, Huajun Chen
| Challenge: | Molecular structure elucidation involves deducing a molecule’s structure from various types of spectral data, which is crucial in chemical experimental analysis. |
| Approach: | They propose a Knowledge-enhanced reasoning framework for Molecular Structure Elucidation that leverages Monte Carlo Tree Search for test-time scaling as a plugin to extend the LLMs’ coverage of the chemical structure space. |
| Outcome: | The proposed framework significantly improves on both GPT-4o-mini and GPT4o, and a specialized molecule-spectrum scorer improves performance. |
MEMERAG: A Multilingual End-to-End Meta-Evaluation Benchmark for Retrieval Augmented Generation (2025.acl-long)
Copied to clipboard
María Andrea Cruz Blandón, Jayasimha Talur, Bruno Charron, Dong Liu, Saab Mansour, Marcello Federico
| Challenge: | Existing benchmarks focus on English or use translated data, which fails to capture cultural nuances. |
| Approach: | They propose to use a multilingual end-to-end Meta-Evaluation RAG benchmark MEMERAG to assess accuracy and faithfulness of RAG systems. |
| Outcome: | The proposed benchmark can identify improvements offered by advanced prompting techniques and LLMs. |
The Role of Visual Modality in Multimodal Mathematical Reasoning: Challenges and Insights (2025.acl-long)
Copied to clipboard
Yufang Liu, Yao Du, Tao Ji, Jianing Wang, Yang Liu, Yuanbin Wu, Aimin Zhou, Mengdi Zhang, Xunliang Cai
| Challenge: | Existing models that leverage visual information do not improve math reasoning performance . authors suggest that visual information is important for multimodal reasoning . |
| Approach: | They propose a dataset to require image reliance for problem-solving and challenge models with similar, yet distinct, images that change the correct answer. |
| Outcome: | The proposed model performance is unaffected by changes to or removal of images in the dataset. |
The Essence of Contextual Understanding in Theory of Mind: A Study on Question Answering with Story Characters (2025.acl-long)
Copied to clipboard
Chulun Zhou, Qiujing Wang, Mo Yu, Xiaoqian Yue, Rui Lu, Jiangnan Li, Yifan Zhou, Shunchi Zhang, Jie Zhou, Wai Lam
| Challenge: | Theory-of-Mind (ToM) is a psychological capability that allows humans to understand and interpret the mental states of others. |
| Approach: | They propose a CharToM-QA benchmark to assess the importance of comprehensive contextual understanding about personal backgrounds in ToM. |
| Outcome: | The proposed model outperforms existing models on 1,035 ToM questions based on classic novels and shows that educated participants perform better when they have read the novels than non-educated participants. |
S2R: Teaching LLMs to Self-verify and Self-correct via Reinforcement Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to incentivize LLMs’ deep thinking abilities require large-scale data or significant training efforts. |
| Approach: | They introduce an efficient framework that enhances LLM reasoning by teaching models to self-verify and self-correct during inference. |
| Outcome: | The proposed framework outperforms models trained on long-CoT distilled data with 3.1k initialization samples and achieves an accuracy improvement of 51.0% to 81.6%. |
Advancing Collaborative Debates with Role Differentiation through Multi-Agent Reinforcement Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-agent collaborative tasks exhibit exceptional capabilities in natural language applications and generation. |
| Approach: | They propose a multi-LLM Cooperation framework with automatic role assignment capabilities that allows multiple agents to embed roles in turn-based speaking. |
| Outcome: | The proposed framework improves collaboration and expertise among agents and teams by enabling them to share roles and develop complementary strengths from the optimization level. |
Retrieval-Augmented Fine-Tuning With Preference Optimization For Visual Program Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Visual programming languages (VPLs) allow users to create programs through graphical interfaces, which results in easier accessibility and widespread usage in various domains. |
| Approach: | They propose to train VPLs from user instructions using large language models . they propose to use retrieval-augmented fine-tuning to leverage repetitive use of subroutines . |
| Outcome: | The proposed method outperforms prompting-based methods for LD generation accuracy even with smaller backbone models. |
STRICTA: Structured Reasoning in Critical Text Assessment for Peer Review and Beyond (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work treats critical text assessment as a black box problem, limiting interpretability and human-AI collaboration. |
| Approach: | They propose a framework to model critical text assessment as an explicit, step-wise reasoning process. |
| Outcome: | The proposed framework breaks down assessment into a graph of interconnected reasoning steps drawing on causality theory. |
XDAC: XAI-Driven Detection and Attribution of LLM-Generated News Comments in Korean (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models generate human-like text, raising concerns about their misuse in creating deceptive content. |
| Approach: | They propose a framework for detecting LLM-generated comments in Korean news and introduce a XDAC framework that leverages explainable AI to uncover distinguishing linguistic patterns at token and character levels. |
| Outcome: | The proposed framework outperforms existing methods and achieves 98.5% F1 score in detection and 84.3% F1 in attribution. |
CENTAUR: Bridging the Impossible Trinity of Privacy, Efficiency, and Performance in Privacy-Preserving Transformer Inference (2025.acl-long)
Copied to clipboard
Jinglong Luo, Guanzhong Chen, Yehong Zhang, Shiyu Liu, Hui Wang, Yue Yu, Xun Zhou, Yuan Qi, Zenglin Xu
| Challenge: | Existing privacy-preserving Transformer Inference frameworks suffer from high computational overhead and performance losses. |
| Approach: | They propose a framework that integrates random permutations and SMPC to address the "impossible trinity" CENTAUR resists diverse data reconstruction attacks and boosts inference speed by 5.030.4 times . |
| Outcome: | CENTAUR achieves an unprecedented balance between privacy, efficiency, and performance. |
Silencing Empowerment, Allowing Bigotry: Auditing the Moderation of Hate Speech on Twitch (2025.acl-long)
Copied to clipboard
| Challenge: | To meet the demands of content moderation, online platforms have resorted to automated systems. |
| Approach: | They conduct an audit of Twitch’s automated moderation tool (AutoMod) to investigate its effectiveness in flagging hateful content. |
| Outcome: | The automated moderation tool (AutoMod) is used to filter hateful content on Twitch and send 107,000 comments from 4 datasets. |
EdiText: Controllable Coarse-to-Fine Text Editing with Diffusion Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for text editing have been proposed for various types of data with diverse attributes. |
| Approach: | They propose a novel text editing method that modifies reference text to desired attributes at various scales. |
| Outcome: | The proposed method is capable of making precise adjustments within the desired range while maintaining the accuracy of the reference text. |
TUMLU: A Unified and Native Language Understanding Benchmark for Turkic Languages (2025.acl-long)
Copied to clipboard
Jafar Isbarov, Arofat Akhundjanova, Mammad Hajili, Kavsar Huseynova, Dmitry Gaynullin, Anar Rzayev, Osman Tursun, Aizirek Turdubaeva, Ilshat Saetov, Rinat Kharisov, Saule Belginova, Ariana Kenbayeva, Amina Alisheva, Abdullatif Köksal, Samir Rustamov, Duygu Ataman
| Challenge: | preparing native language MMLU benchmarks is costly and limits representativeness of evaluation datasets. |
| Approach: | They propose to use a Turkic language MMLU benchmark to assess massive multitask language understanding capabilities. |
| Outcome: | The proposed benchmarks are based on a Turkic language morphosyntactic and cultural benchmark . the benchmarks evaluate a diverse range of open and proprietary multilingual large language models . |
Look Both Ways and No Sink: Converting LLMs into Text Encoders without Training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for converting large language models into powerful text encoders require extensive training on large datasets. |
| Approach: | They propose a training-free approach that enables bidirectional attention and suppresses the attention sink phenomenon, resulting in superior performance. |
| Outcome: | The proposed approach enables bidirectional attention and suppresses the attention sink phenomenon, resulting in superior performance. |
A Statistical and Multi-Perspective Revisiting of the Membership Inference Attack in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Membership Inference Attack (MIA) is a method that differentiates trained (member) and untrained (non-member) data. |
| Approach: | They used thousands of experiments to examine membership inference attacks from different settings and then revisited them with thousands of different methods. |
| Outcome: | The proposed methods outperform baselines in the study and improve with model size and varies with domains. |
Around the World in 24 Hours: Probing LLM Knowledge of Time and Place (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have tested language models' ability to reason over time and space in isolation or only in simple or artificial environments. |
| Approach: | They present a dataset of 320k prompts covering 289 cities in 217 countries and 37 time zones to evaluate their ability to jointly reason over time and space. |
| Outcome: | The proposed models perform well on reasoning tasks involving only temporal knowledge, but performance remains constrained on tasks that require connecting temporal and geographic information. |
Mining the uncertainty patterns of humans and models in the annotation of moral foundations and human values (2025.acl-long)
Copied to clipboard
| Challenge: | disagreement in annotation (HLV) is considered a constitutive feature of subjective tasks. |
| Approach: | They investigate the relationship between disagreement in annotation and model uncertainty . they use linguistic features to calibrate models to HLV and uncertainty to analyze their impact on uncertainty. |
| Outcome: | The proposed model uncertainty is calibrated to human label variation (HLV) the proposed model is calibrate to human labels, the authors show . |
“What do you call a dog that is incontrovertibly true? Dogma”: Testing LLM Generalization through Humor (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have shown strong performance in NLP tasks like text summarization and question answering. |
| Approach: | They propose a new humor-based question-answering benchmark to assess LLMs’ reasoning through carefully crafted puns. |
| Outcome: | Experiments on pun comprehension, resolution, and generation reveal that most LLMs struggle with generalization, even on simple tasks, consistently underperforming the human baseline. |
Towards Harmonized Uncertainty Estimation for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have demonstrated exceptional capabilities in handling a wide range of downstream tasks. |
| Approach: | They propose a method that employs a lightweight model trained on data aligned with the target LLM’s performance to adjust uncertainty scores. |
| Outcome: | The proposed method achieves improvements of up to 60% over existing methods. |
VITAL: A New Dataset for Benchmarking Pluralistic Alignment in Healthcare (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to align Large Language Models with human values model an averaged or monolithic preference, despite progress in pluralistic alignment, no prior work has focused on health . |
| Approach: | They propose a benchmark dataset to assess and benchmark pluralistic alignment methodologies. |
| Outcome: | The proposed model can model pluralistic views within health domains. |
Are We in the AI-Generated Text World Already? Quantifying and Monitoring AIGT on Social Media (2025.acl-long)
Copied to clipboard
| Challenge: | Social media platforms are experiencing a growing presence of AI-Generated Texts (AIGTs) however, the misuse of AIGTs could have profound implications for public opinion . |
| Approach: | They collect a dataset with 2.4M posts from 3 major social media platforms . they then construct a diverse dataset to train and evaluate AIGT detectors . |
| Outcome: | The proposed dataset analyzes 2.4M posts from 3 major social media platforms from 2022 to 2024 . it finds that Medium and Quora show marked increases in AAR . |
From English to Second Language Mastery: Enhancing LLMs with Cross-Lingual Continued Instruction Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) acquire strong language skills through extensive pre-training and supervised fine-tuning (SFT) on instructionresponse pairs. |
| Approach: | They propose a method which leverages translation-based parallel instruction data to enhance cross-lingual adaptability. |
| Outcome: | The proposed model improves on Llama-2-7B across five languages against three objective benchmarks and an LLM-as-a-judge benchmark. |
WET: Overcoming Paraphrasing Vulnerabilities in Embeddings-as-a-Service with Linear Transformation Watermarks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing EaaS watermarks can be removed by paraphrasing when attackers clone the model. |
| Approach: | They propose a method that integrates a target embedding into the original embeddable based on the presence of trigger words in the input text. |
| Outcome: | The proposed technique is empirically and theoretically robust against paraphrasing. |
HoPE: A Novel Positional Encoding Without Long-Term Decay for Enhanced Context Awareness and Extrapolation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing positional encodings exhibit long-term decay, based on an entrenched and long-standing opinion that tokens farther away from the current position carry less relevant information. |
| Approach: | They propose a high-frequency rotary position encoding (HoPE) that replaces specific components in RoPE with position-independent ones, retaining only high- frequency signals. |
| Outcome: | The proposed method exhibits greater robustness to the out-of-distribution behavior in attention patterns during extrapolation. |
One QuantLLM for ALL: Fine-tuning Quantized LLMs Once for Efficient Deployments (2025.acl-long)
Copied to clipboard
| Challenge: | Quantization has shown promise for Large Language Models, but current methods require lengthy training to alleviate quantization loss. |
| Approach: | They propose to decouple weights and incorporate Low-Rank adapters to reduce weight sharing . they validate the approach on LLaMA2 families and Mistral on downstream evaluation . |
| Outcome: | The proposed approach shows high performance while reducing deployment time faced with multiple scenarios. |
Beyond Logits: Aligning Feature Dynamics for Effective Knowledge Distillation (2025.acl-long)
Copied to clipboard
Guoqiang Gong, Jiaxing Wang, Jin Xu, Deping Xiang, Zicheng Zhang, Leqi Shen, Yifeng Zhang, JunhuaShu JunhuaShu, ZhaolongXing ZhaolongXing, Zhen Chen, Pengzhang Liu, Ke Zhang
| Challenge: | Knowledge distillation (KD) compresses large language models into lightweight versions called student models. |
| Approach: | They propose to align the entire feature dynamics between teacher and student models by using two additional loss terms to achieve this. |
| Outcome: | The proposed method matches the entire feature dynamics between teacher and student models rather than just the final states. |
Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention (2025.acl-long)
Copied to clipboard
Jingyang Yuan, Huazuo Gao, Damai Dai, Junyu Luo, Liang Zhao, Zhengyan Zhang, Zhenda Xie, Yuxing Wei, Lean Wang, Zhiping Xiao, Yuqing Wang, Chong Ruan, Ming Zhang, Wenfeng Liang, Wangding Zeng
| Challenge: | Long-context modeling is crucial for next-generation language models, but high computational cost of standard attention mechanisms poses significant computational challenges. |
| Approach: | They propose a natively trained Sparse Attention mechanism that integrates algorithms with hardware-aligned optimizations to achieve efficient long-context modeling. |
| Outcome: | The proposed model maintains or exceeds Full Attention models across general benchmarks, long-context tasks, and instruction-based reasoning. |
DRAE: Dynamic Retrieval-Augmented Expert Networks for Lifelong Learning and Task Adaptation in Robotics (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that Dynamic Retrieval-Augmented Expert Networks outperforms baseline approaches in long-term task retention and knowledge reuse. |
| Approach: | They propose a dynamic routing architecture that leverages MoE and Retrieval-Augmented Generation to augment the learning process. |
| Outcome: | The proposed architecture outperforms baseline approaches in long-term task retention and knowledge reuse. |
MT-RAIG: Novel Benchmark and Evaluation Framework for Retrieval-Augmented Insight Generation over Multiple Tables (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on table-based reasoning focus on a single gold table, not multiple tables . a persistent demand for robust table understanding systems is resulting from the complexity of table data . |
| Approach: | They propose a MT-RAIG Bench to evaluate systems on Retrieval-Augmented Insight Generation over Mulit-Tables. |
| Outcome: | The proposed framework improves human quality judgments on the generated insights. |
Enhancing Chain-of-Thought Reasoning with Critical Representation Fine-tuning (2025.acl-long)
Copied to clipboard
Chenxi Huang, Shaotian Yan, Liang Xie, Binbin Lin, Sinan Fan, Yue Xin, Deng Cai, Chen Shen, Jieping Ye
| Challenge: | Representation Fine-tuning (ReFT) is a proposed method for improving parameter efficiency . however, it yields suboptimal performance, as fixed-position representations have uncertain impact on outputs . |
| Approach: | They propose a method that fine-tunes critical representations in a low-rank linear subspace while freezing the base model. |
| Outcome: | The proposed method improves accuracy of LLaMA-2-7B and ReFT by 18.2 and 3.8 on GSM8K. |
Does the Emotional Understanding of LVLMs Vary Under High-Stress Environments and Across Different Demographic Attributes? (2025.acl-long)
Copied to clipboard
| Challenge: | According to psychological and neuroscientific research, a high-stress environment can restrict attentional resources and intensify negative affect, thereby impairing the ability to understand emotions. |
| Approach: | They constructed a large-vision language model that combines race, gender, and age group and used the Pretend prompt technique to induce LVLMs to interpret others’ emotions. |
| Outcome: | The results suggest that the effects of high-stress and demographic attributes identified in human research may also be reflected in LVLMs. |
S2WTM: Spherical Sliced-Wasserstein Autoencoder for Topic Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Variational autoencoder-based neural topic models lack posterior collapse in high-dimensional text data. |
| Approach: | They propose a Spherical Sliced Wasserstein Autoencoder for Topic Modeling to model latent representations in a hyperspherical space. |
| Outcome: | The proposed model outperforms state-of-the-art topic models while improving performance on downstream tasks. |
Learning to Look at the Other Side: A Semantic Probing Study of Word Embeddings in LLMs with Enabled Bidirectional Attention (2025.acl-long)
Copied to clipboard
| Challenge: | Autoregressive Large Language Models (LLMs) demonstrate exceptional performance in language understanding and generation tasks, but their application in text embedding tasks has been relatively slow due to the constraints of the unidirectional attention mechanism. |
| Approach: | They propose to enable bidirectional attention in LLMs to improve their ability to represent subsequent context but weaken their utilization of preceding context. |
| Outcome: | The proposed model improves the LLMs’ ability to represent subsequent context but weakens their utilization of preceding context, while contrastive learning training can help to maintain both abilities. |
Tracing and Dissecting How LLMs Recall Factual Knowledge for Real World Questions (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models have shown promising ability to perform commonsense reasoning. |
| Approach: | They propose a two-dimensional analysis framework that incorporates token back-tracing and token decoding to uncover how LLMs conduct factual knowledge recall. |
| Outcome: | The proposed framework shows that LLMs lack relevant knowledge but struggle to select the most accurate information based on context during the retrieval and rerank phase. |
Employing Discourse Coherence Enhancement to Improve Cross-Document Event and Entity Coreference Resolution (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work on cross-document coreference resolution focuses on within-document events and entities, but cross-doc mentions lack such critical contexts. |
| Approach: | They propose a task to enhance the discourse coherence between two cross-document mentions by adding coherent texts to a document to form a new coherent document. |
| Outcome: | The proposed method outperforms state-of-the-art baselines on three popular datasets. |
Data Whisperer: Efficient Data Selection for Task-Specific LLM Fine-Tuning via Few-Shot In-Context Learning (2025.acl-long)
Copied to clipboard
Shaobo Wang, Xiangqi Jin, Ziming Wang, Jize Wang, Jiajun Zhang, Kaixin Li, Zichen Wen, Zhong Li, Conghui He, Xuming Hu, Linfeng Zhang
| Challenge: | Using fine-tuning on task-specific data is essential for large language models to be effective in specialized tasks. |
| Approach: | They propose a method that leverages few-shot in-context learning with the model to be fine-tuned. |
| Outcome: | The proposed method outperforms existing methods with a 3.1-point improvement and a 7.4 speedup on the Llama-3-8B-Instruct model using just 10% of the dataset. |
Synthesizing Post-Training Data for LLMs through Multi-Agent Simulation (2025.acl-long)
Copied to clipboard
Shuo Tang, Xianghe Pang, Zexi Liu, Bohan Tang, Rui Ye, Tian Jin, Xiaowen Dong, Yanfeng Wang, Siheng Chen
| Challenge: | Extensive experiments demonstrate that our framework effectively generates both general and domain-specific data. |
| Approach: | They propose a multi-agent simulator that automatically generates diverse text-based scenarios, capturing a wide range of real-world human needs. |
| Outcome: | Experiments show that the proposed model outperforms Meta’s Llama-3-8B-Instruct model on AlpacaEval 2 and Arena-Hard benchmarks with just 20K instruction-response pairs. |
SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to continuous-space reasoning focus on hard token decoding and suffer from catastrophic forgetting. |
| Approach: | They propose a method that generates instance-specific soft thought tokens as the initial chain of thoughts and maps them into the LLM’s representation space via a trainable projection module. |
| Outcome: | The proposed method improves LLM reasoning performance through supervised, parameter-efficient fine-tuning. |
FCMR: Robust Evaluation of Financial Cross-Modal Multi-Hop Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks for multimodal large language models lack data contamination and complex queries . financial cross-modal multi-hop reasoning is difficult to evaluate and requires precise cross-module reasoning . |
| Approach: | They propose a benchmark to analyze the reasoning capabilities of multimodal large language models. |
| Outcome: | The proposed model is categorized into three difficulty levels—easy, medium, and hard—for step-by-step evaluation. |
Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms (2025.acl-long)
Copied to clipboard
| Challenge: | Recent research has explored the use of sparse autoencoders (SAE) to disentangle knowledge in high-dimensional spaces for steering. |
| Approach: | They propose a method that isolates and manipulates disentangled knowledge components to enhance safety by using sparse autoencoders to disentangle knowledge in high-dimensional spaces for steering. |
| Outcome: | The proposed method is able to isolate and manipulate disentangled knowledge components to enhance safety in large reasoning models. |
MobiLoRA: Accelerating LoRA-based LLM Inference on Mobile Devices via Context-aware KV Cache Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | MobiLoRA focuses on optimizing the key-value (KV) caches due to the limited computing and memory resources of mobile devices. |
| Approach: | They propose to optimize the key-value caches due to limited computing resources . they propose similarity-aware delta encoding for semantic-level contexts . |
| Outcome: | The proposed model accelerates LoRA-based LLM inference by 57.6% on mobile devices. |
Language Models Resist Alignment: Evidence From Data Compression (2025.acl-long)
Copied to clipboard
Jiaming Ji, Kaile Wang, Tianyi Alex Qiu, Boyuan Chen, Jiayi Zhou, Changye Li, Hantao Lou, Josef Dai, Yunhuai Liu, Yaodong Yang
| Challenge: | Large language models (LLMs) may exhibit undesirable behaviors due to the inevitable biases and harmful content present in training. |
| Approach: | They propose to investigate the elasticity of large language models by examining their performance. |
| Outcome: | The proposed model performance declines rapidly before reverting to the pre-training distribution, the authors show . the proposed model weight and code are available at pku-lm-res ist-alignment.github.io. |
Beyond the Answer: Advancing Multi-Hop QA with Fine-Grained Graph Reasoning and Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluations of multi-hop question answering systems focus on comparing final answers of reasoning method and given ground-truths. |
| Approach: | They propose a "Planner-Executor-Reasoner" architecture that evaluates reasoning . they propose PER-DP and PER QA architectures that provide ground-truths . |
| Outcome: | The proposed model improves the performance of multi-hop question answering systems. |
Mamba Knockout for Unraveling Factual Information Flow (2025.acl-long)
Copied to clipboard
| Challenge: | Recent work has introduced Mamba-based SSM architectures that rival Transformer performance in various settings. |
| Approach: | They propose to use attentional interpretability techniques originally developed for Transformers to trace how information is transmitted and localized across tokens and layers. |
| Outcome: | The proposed model disentangles how distinct features enable token-to-token information exchange or enrich individual tokens, thus offering a unified lens to understand Mamba internal operations. |
Small Changes, Big Impact: How Manipulating a Few Neurons Can Drastically Alter LLM Aggression (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models have led to innovations in various domains such as education, healthcare, and finance, while raising serious concerns that they can be easily misused for malicious purposes. |
| Approach: | They identify specific neurons (“aggression neurons”) closely related to the expression of aggression and analyze how manipulating them affects the model’s overall aggression. |
| Outcome: | The proposed model outputs show that manipulating neurons can increase aggression by up to 33% in all models and even more extreme when they are concentrated in certain layers. |
Marco-o1 v2: Towards Widening The Distillation Bottleneck for Reasoning Models (2025.acl-long)
Copied to clipboard
Huifeng Yin, Yu Zhao, Minghao Wu, Xuanfan Ni, Bo Zeng, Huaiyu.wh Huaiyu.wh, Tianqi Shi, Liangying Shao, Chenyang Lyu, Longyue Wang, Weihua Luo, Kaifu Zhang
| Challenge: | Recent efforts to distill large reasoning models into smaller lightweight models have shown competitive performances. |
| Approach: | They propose to distill long Chain-of-Thought data to improve SFT and RL methods by constructing data from scratch using Monte Carlo Tree Search. |
| Outcome: | The proposed method significantly improves reasoning performance on various benchmarks such as math (GSM8K, MATH, AIME). |
Curiosity-Driven Reinforcement Learning from Human Feedback (2025.acl-long)
Copied to clipboard
| Challenge: | Reinforcement learning from human feedback (RLHF) has proven effective in aligning large language models with human preferences, but often at the cost of reduced output diversity. |
| Approach: | They propose a framework that incorporates intrinsic rewards for novel states alongside traditional sparse extrinsic rewards to optimize both output diversity and alignment quality. |
| Outcome: | The proposed framework achieves significant gains in diversity on multiple diversity-oriented metrics while maintaining alignment with human preferences comparable to standard RLHF. |
T2A-Feedback: Improving Basic Capabilities of Text-to-Audio Generation via Fine-grained AI Feedback (2025.acl-long)
Copied to clipboard
Zehan Wang, Ke Lei, Chen Zhu, Jiawei Huang, Sashuai Zhou, Luping Liu, Xize Cheng, Shengpeng Ji, Zhenhui Ye, Tao Jin, Zhou Zhao
| Challenge: | Text-to-audio (T2A) models still struggle to satisfy human preferences for prompt-following and acoustic quality when generating complex multi-event audio. |
| Approach: | They propose to use AI feedback learning to enhance basic capabilities of text-to-audio models . they use a large audio preference dataset to evaluate the model's capabilities . |
| Outcome: | The proposed model improves in simple and complex scenarios with AI feedback learning. |
CoE: A Clue of Emotion Framework for Emotion Recognition in Conversations (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are limited in interpreting complex conversational streams. |
| Approach: | They propose a Clue of Emotion framework which integrates key conversational clues to enhance the ERC task. |
| Outcome: | The proposed framework outperforms EmoryNLP, MELD, and IEMOCAP in the role-playing, speaker identification, and emotion reasoning tasks. |
MPO: Multilingual Safety Alignment via Reward Gap Optimization (2025.acl-long)
Copied to clipboard
Weixiang Zhao, Yulin Hu, Yang Deng, Tongtong Wu, Wenxuan Zhang, Jiahe Guo, An Zhang, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu
| Challenge: | Existing preference learning methods for safety alignment are monolingual and struggle with noisy multilingual data. |
| Approach: | They propose a multilingual reward gaP optimization approach that leverages the well-aligned safety capabilities of the dominant language to improve safety alignment across multiple languages. |
| Outcome: | Extensive experiments on three LLMs, LLaMA-3.1, Gemma-2 and Qwen2.5, validate MPO’s efficacy in multilingual safety alignment without degrading general multilingual utility. |
QualiSpeech: A Speech Quality Assessment Dataset with Natural Language Reasoning and Descriptions (2025.acl-long)
Copied to clipboard
Siyin Wang, Wenyi Yu, Xianzhao Chen, Xiaohai Tian, Jun Zhang, Lu Lu, Yu Tsao, Junichi Yamagishi, Yuxuan Wang, Chao Zhang
| Challenge: | Existing datasets lack comprehensive annotations for speech quality assessment . existing methods lack detailed annotations, resulting in inaccurate evaluations. |
| Approach: | They propose a low-level speech quality assessment dataset incorporating natural language descriptions and a Benchmark to evaluate low- level speech understanding capabilities of auditory large language models. |
| Outcome: | The proposed model can be used to evaluate the low-level speech understanding capabilities of auditory large language models. |
On the Relation Between Fine-Tuning, Topological Properties, and Task Performance in Sense-Enhanced Embeddings (2025.acl-long)
Copied to clipboard
| Challenge: | Enhanced word embeddings do not align well with word senses, resulting in poor performance on word sense identification tasks. |
| Approach: | They propose to use two methods to fine-tune embeddings to identify the topological properties that contribute to sense-enhanced embeddables. |
| Outcome: | The proposed methods improve the embeddings’ ability to capture nuanced semantic distinctions while reducing their expressiveness. |
Finding Needles in Images: Can Multi-modal LLMs Locate Fine Details? (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Multi-modal Large Language Models (MLLMs) have fundamentally transformed how machines understand and reason about visual information. |
| Approach: | They propose a benchmark to evaluate MLLMs' ability to locate and reason about fine-grained details within complex documents including newspapers, menus, and lecture images. |
| Outcome: | The proposed method improves on existing methods and shows that it can handle fine-grained document understanding tasks. |
Don’t Half-listen: Capturing Key-part Information in Continual Instruction Tuning (2025.acl-long)
Copied to clipboard
Yongquan He, Wenyuan Zhang, Xuancheng Huang, Peng Zhang, Lingxun Meng, Xiang Zhou, Ke Zeng, Xunliang Cai
| Challenge: | Existing methods to improve instruction tuning for large language models may cause catastrophic forgetting (CF) CF is a problem where previously learned abilities are degraded . |
| Approach: | They propose a continual instruction tuning method that uses key-part information gain to replay data and refine training objective. |
| Outcome: | The proposed method achieves superior performance on both seen and held-out tasks. |
Generating Plausible Distractors for Multiple-Choice Questions via Student Choice Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Multiple-choice questions (MCQs) are critical for identifying misconceptions and gaps in knowledge and accurately assessing students' understanding. |
| Approach: | They propose to train a model to generate distractors that are more likely to be selected by students by a pairwise ranker and a distractor generator via Direct Preference Optimization. |
| Outcome: | The proposed model outperforms baseline models and performs comparable to humans in various metrics including pairwise rank accuracy and distractor plausibility. |
Exploring Explanations Improves the Robustness of In-Context Learning (2025.acl-long)
Copied to clipboard
| Challenge: | In-context learning (ICL) has been shown to be effective across a variety of tasks, but it has been reported to be restricted in its ability to generalize beyond the given demonstrations. |
| Approach: | They propose a framework that extends ICL by exploring explanations for all possible labels. |
| Outcome: | The proposed framework improves prediction reliability by exploring explanations for all possible labels. |
Prediction Hubs are Context-Informed Frequent Tokens in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Hubness is a tendency for a few points to be among the nearest neighbours of a disproportionate number of other points. |
| Approach: | They show that only large-scale representation comparisons are not characterized by hubness . they show that hubs are the result of context-modulated frequent tokens . |
| Outcome: | The results show that the comparison between context and unembedding vectors does not result in hubness . the findings suggest that hubness is not a negative property that needs to be mitigated when LLMs are being used for next token prediction. |
Capability Salience Vector: Fine-grained Alignment of Loss and Capabilities for Downstream Task Scaling Law (2025.acl-long)
Copied to clipboard
Qiming Ge, Shuhao Xing, Songyang Gao, Yunhua Zhou, Yicheng Zou, Songyang Zhang, Zhi Chen, Hang Yan, Qi Zhang, Qipeng Guo, Kai Chen
| Challenge: | Large language models have demonstrated impressive performance across a wide range of tasks, but this achievement comes with the trade-off of significant computational demands. |
| Approach: | They propose a scaling law that decomposes the overall validation loss and assigns different importance weights to tokens to assess a specific meta-capability. |
| Outcome: | The proposed model can predict the loss trending of models across different levels of computation without a gap between validation loss and model's downstream capabilities. |
CRUXEVAL-X: A Benchmark for Multilingual Code Reasoning, Understanding and Execution (2025.acl-long)
Copied to clipboard
Ruiyang Xu, Jialun Cao, Yaojie Lu, Ming Wen, Hongyu Lin, Xianpei Han, Ben He, Shing-Chi Cheung, Le Sun
| Challenge: | Existing code benchmarks focus on code generation, while those for code reasoning are insufficient. |
| Approach: | They propose a multi-lingual code reasoning benchmark that contains 19 programming languages and at least 600 subjects for each language. |
| Outcome: | The proposed model trains on Python and achieves 34.4% Pass@1 in other languages, revealing the cross-language generalization of LLMs. |
Graph of Records: Boosting Retrieval Augmented Generation for Long-context Summarization with Graphs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for retrieving historical LLM responses are lacking in long-context summarization tasks. |
| Approach: | They propose a graph of records which leverages historical LLM responses to enhance RAG for long-context global summarization. |
| Outcome: | The proposed method improves on four long-context summarization datasets. |
Rubrik’s Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset (2025.acl-long)
Copied to clipboard
Diana Galvan-Sosa, Gabrielle Gaudeau, Pride Kavumba, Yunmeng Li, Hongyi Gu, Zheng Yuan, Keisuke Sakaguchi, Paula Buttery
| Challenge: | Large-Language Models (LLMs) are increasingly being used in explanation generation tasks due to their unreliability. |
| Approach: | They propose a rubric and a dataset of 26k explanations written and quality-annotated using the rubric by humans and six open- and closed-source LLMs to test their proposed rubric. |
| Outcome: | The proposed rubric and CUBE dataset focuses on reasoning and language tasks and provides the necessary diversity to test it. |
A Dual-Mind Framework for Strategic and Expressive Negotiation Agent (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to negotiation dialogue focus on only one aspect, ignoring the synergistic effect of their combined synergies. |
| Approach: | They propose a dual-mind negotiation agent framework that integrates an intuitive and a deliberative module for slow, expression optimization. |
| Outcome: | The proposed framework achieves state-of-the-art on negotiation datasets showing that it improves negotiation ability. |
Ref-Long: Benchmarking the Long-context Referencing Capability of Long-context Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Long-context language models have impressive capabilities in long-contrast understanding tasks, but long-text referencing remains underexplored. |
| Approach: | They propose a benchmark to assess long-context referencing capability of LCLMs . they use three subsets to test the model's ability to identify key indexes based on contextual relationships . |
| Outcome: | The proposed benchmark assesses the long-context referencing capability of LCLMs. |
Revisiting Scaling Laws for Language Models: The Role of Data Quality and Training Strategies (2025.acl-long)
Copied to clipboard
| Challenge: | Existing scaling laws suggest augmenting model size and training data results in enhanced performance, but recent studies reveal deviations, particularly in large language models, where performance improvements decelerate—a phenomenon known as sub-scaling. |
| Approach: | They propose a sub-optimal scaling law that better predicts performance in sub-scaling regimes by examining data quality and training strategies. |
| Outcome: | The proposed scaling law better predicts performance in sub-scaling regimes, highlighting the importance of data quality and diversity. |
Limited Generalizability in Argument Mining: State-Of-The-Art Models Learn Datasets, Not Arguments (2025.acl-long)
Copied to clipboard
| Challenge: | Identifying arguments is a prerequisite for various tasks in automated discourse analysis. |
| Approach: | They evaluate four BERT-like transformers on 17 English sentence-level datasets . they find that they tend to rely on lexical shortcuts tied to content words . |
| Outcome: | The proposed models perform best on 17 English sentence-level datasets on common tasks, but their performance drops when applied to unseen datasets. |
Enhancing Machine Translation with Self-Supervised Preference Data (2025.acl-long)
Copied to clipboard
| Challenge: | Current approaches to constructing preference data rely on human annotations. |
| Approach: | They propose a framework which efficiently constructs translation preference data for iterative training. |
| Outcome: | The proposed framework improves translation preference data on large language models. |
Unveil: Unified Visual-Textual Integration and Distillation for Multi-modal Document Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Document retrieval in real-world scenarios faces significant challenges due to diverse document formats and modalities. |
| Approach: | They propose a visual-textual embedding framework that integrates textual and visual features for robust document representation. |
| Outcome: | The proposed visual-textual embedding framework surpasses existing methods while preserving semantic fidelity. |
Don’t Get Lost in the Trees: Streamlining LLM Reasoning by Overcoming Tree Search Exploration Pitfalls (2025.acl-long)
Copied to clipboard
Ante Wang, Linfeng Song, Ye Tian, Dian Yu, Haitao Mi, Xiangyu Duan, Zhaopeng Tu, Jinsong Su, Dong Yu
| Challenge: | Recent advances in tree search algorithms guided by verifiers have significantly enhanced the reasoning capabilities of large language models (LLMs), but at the cost of increased computational resources. |
| Approach: | They propose an e ffici ent tree sear ch framework that is a plug-and-play system compatible with various tree search algorithms. |
| Outcome: | The proposed framework reduces computational costs and prioritizes resource allocation to harder tasks (Levels 3-4) over simpler ones (Level 1-2), addressing both over-exploration in basic problems and under-exploation in complex cases. |
MEXMA: Token-level objectives improve sentence representations (2025.acl-long)
Copied to clipboard
| Challenge: | Current approaches to cross-lingual sentence encoders use sentence-level objectives only. |
| Approach: | They propose a novel approach that integrates both sentence-level and token-level objectives. |
| Outcome: | The proposed approach outperforms existing CLSEs on bitext mining tasks and downstream tasks. |
Uncertainty-Aware Iterative Preference Optimization for Enhanced LLM Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for enhancing the performance of large language models require expensive manual annotations. |
| Approach: | They propose an offline direct preference optimization method that collects preference pairs through iterative sampling and execution feedback to improve model confidence. |
| Outcome: | The proposed method improves performance on three reasoning tasks and shows a 3.6% improvement over the standard method. |
AgentDropout: Dynamic Agent Elimination for Token-Efficient and High-Performance LLM-Based Multi-Agent Collaboration (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for MAS suffer from high token consumption and inefficiency due to frequent generation and communication among multiple agents. |
| Approach: | They propose a multi-agent system based on large language models that identifies redundant agents and communication across different communication rounds by optimizing the adjacency matrices of the communication graphs and eliminates them to enhance both token efficiency and task performance. |
| Outcome: | The proposed method reduces prompt token consumption and completion token consumption by 18.4% and improves task performance by 1.14. |
Towards Dynamic Theory of Mind: Evaluating LLM Adaptation to Temporal Evolution of Human States (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks assess basic Theory of Mind abilities but neglect temporal evolution of mental states in real-world social contexts. |
| Approach: | They propose a benchmark specifically designed to evaluate Large Language Models' ability to understand and track the temporal progression of mental states across interconnected scenarios. |
| Outcome: | The proposed benchmarks underperform humans by 44.7% and show that they can model the dynamic nature of human mental states better than existing models. |
Marco-Bench-MIF: On Multilingual Instruction-Following Capability of Large Language (2025.acl-long)
Copied to clipboard
Bo Zeng, Chenyang Lyu, Sinuo Liu, Mingyan Zeng, Minghao Wu, Xuanfan Ni, Tianqi Shi, Yu Zhao, Yefeng Liu, Chenyu Zhu, Ruizhe Li, Jiahui Geng, Qing Li, Yu Tong, Longyue Wang, Weihua Luo, Kaifu Zhang
| Challenge: | Existing datasets for instruction-following are monolingual and centered on English . existing data are unable to capture linguistic and cultural subtle differences . |
| Approach: | They propose an extension of IFEval to a localized multilingual version called Marco-Bench-MIF . their benchmark addresses linguistic constraints and cultural references via translation and verification . |
| Outcome: | The proposed extension of IFEval to a localized multilingual version covers 30 languages with varying levels of localization. |
Representation Bending for Large Language Model Safety (2025.acl-long)
Copied to clipboard
Ashkan Yousefpour, Taeheon Kim, Ryan Sungmo Kwon, Seungbeen Lee, Wonje Jeung, Seungju Han, Alvin Wan, Harrison Ngan, Youngjae Yu, Jonghyun Choi
| Challenge: | Existing safety-enhancing techniques, such as fine-tuning with human feedback or adversarial training, are still vulnerable as they address specific threats and fail to generalize across unseen attacks. |
| Approach: | They propose a new approach that disrupts representations underlying harmful behaviors in Large Language Models by using loss-based fine-tuning. |
| Outcome: | The proposed approach outperforms existing methods such as Circuit Breaker, RMU, and NPO with 95% reduction in attack success rates across diverse jailbreak benchmarks. |
Analyzing LLMs’ Knowledge Boundary Cognition Across Languages Through the Lens of Internal Representations (2025.acl-long)
Copied to clipboard
| Challenge: | Understanding the knowledge boundaries of Large Language Models (LLMs) is crucial to prevent hallucination, but research on the knowledge boundary perceptions of LLMs has predominantly focused on English. |
| Approach: | They propose a training-free alignment method that effectively transfers knowledge boundary perception ability across languages, thereby helping reduce hallucination risk in low-resource languages. |
| Outcome: | The proposed method reduces hallucination risk in low-resource languages by fine-tuning on bilingual question pair translation. |
Enhancing Retrieval-Augmented Generation via Evidence Tree Search (2025.acl-long)
Copied to clipboard
| Challenge: | Evidence retrieval is used to enhance Large Language Models (LLMs) but in real-world applications, it often returns lengthy documents with redundant or irrelevant content, confusing downstream readers. |
| Approach: | They propose a framework that reformulates evidence retrieval as a dynamic tree expansion process. |
| Outcome: | The proposed framework outperforms existing methods on five datasets. |
HalluLens: LLM Hallucination Benchmark (2025.acl-long)
Copied to clipboard
Yejin Bang, Ziwei Ji, Alan Schelten, Anthony Hartshorn, Tara Fowler, Cheng Zhang, Nicola Cancedda, Pascale Fung
| Challenge: | Large language models (LLMs) generate responses that deviate from user input or training data, a phenomenon known as "hallucination" . |
| Approach: | They propose a hallucination benchmark HalluLens that includes both extrinsic and intrinsic evaluation tasks to distinguish between extrindic and intrinsic hallucines. |
| Outcome: | The proposed framework disentangles LLM hallucination from "factuality" and distinguishes between extrinsic and intrinsic hallucines to promote consistency and facilitate research. |
DEEPER Insight into Your User: Directed Persona Refinement for Dynamic Persona Modeling (2025.acl-long)
Copied to clipboard
Aili Chen, Chengyu Du, Jiangjie Chen, Jinghan Xu, Yikai Zhang, Siyu Yuan, Zulong Chen, Liangyue Li, Yanghua Xiao
| Challenge: | Existing methods for generating personas from static historical data fail to capture dynamic behaviors and evolving preferences in real-world interactive scenarios. |
| Approach: | They propose a novel approach that iteratively updates personas using streaming user behavior data to continually enhance their quality. |
| Outcome: | The proposed approach delivers 32.2% reduction in user behavior prediction error over four update rounds, outperforming the best baseline by 22.92%. |
Asclepius: A Spectrum Evaluation Benchmark for Medical Multi-Modal Large Language Models (2025.acl-long)
Copied to clipboard
Jie Liu, Wenxuan Wang, Su Yihang, Jingyuan Huang, Yudi Zhang, Cheng-Yi Li, Wenting Chen, Xiaohan Xing, Kao-Jung Chang, Linlin Shen, Michael R. Lyu
| Challenge: | Medical Multi-Modal Large Language Models (Med-MLLMs) are a promising new form of artificial general intelligence due to their ability to tackle complex tasks. |
| Approach: | They propose a new benchmark that comprehensively assesses medical multi-modal large language models in terms of distinct medical specialties and different diagnostic capacities. |
| Outcome: | The proposed model covers 15 medical specialties and different diagnostic capacities, and excludes overlap with existing VQA dataset. |
InstructPart: Task-Oriented Part Segmentation with Instruction Reasoning (2025.acl-long)
Copied to clipboard
Zifu Wan, Yaqi Xie, Ce Zhang, Zhiqiu Lin, Zihan Wang, Simon Stepputtis, Deva Ramanan, Katia P. Sycara
| Challenge: | Large multimodal foundation models perceive objects as indivisible, overlooking the components that constitute them. |
| Approach: | They propose a novel benchmark for large multimodal foundation models comprising hand-labeled part segmentation annotations and task-oriented instructions to evaluate their performance. |
| Outcome: | The proposed benchmark improves performance of current models in understanding and executing part-level tasks within everyday contexts. |
GRaMPa: Subword Regularisation by Skewing Uniform Segmentation Distributions with an Efficient Path-counting Markov Model (2025.acl-long)
Copied to clipboard
| Challenge: | Subword regularisations are known to be stochastic, but only a handful of possible segmentations are sampled. |
| Approach: | They propose to randomise word segmentations from a subword tokeniser instead of randomising them by weighting paths in an unweighted segmentation graph. |
| Outcome: | The proposed method outperforms existing methods on token-level tasks with spelling errors. |
Evaluating the Evaluation of Diversity in Commonsense Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation metrics for commonsense generation are unclear on which metrics are best suited for evaluating the diversity of outputs. |
| Approach: | They propose to use a large language model to analyze commonsense generation data to determine which diversity metrics are best suited for commonsensing. |
| Outcome: | The proposed metrics outperform form-based metrics and show high correlations with the LLM-based ratings. |
Generate First, Then Sample: Enhancing Fake News Detection with LLM-Augmented Reinforced Sampling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models have a performance gap of 20% between classifying fake news and real news, making them less suitable for practical deployment. |
| Approach: | They propose to adopt an LLM to generate fake news in three different styles, which are later incorporated into the training set to augment the representation of fake news. |
| Outcome: | The proposed model achieves state-of-the-art performance on two benchmark datasets and improves detection accuracy by 24.02% and 11.06% respectively. |
ChemActor: Enhancing Automated Extraction of Chemical Synthesis Actions with LLM-Generated Data (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for extracting chemical procedures from literature are insufficient and low-quality due to the inherent ambiguity of chemical language and the high cost of human annotation. |
| Approach: | They propose a fully fine-tuned large language model (LLM) as a chemical executor to convert between unstructured experimental procedures and structured action sequences. |
| Outcome: | The proposed model outperforms the baseline model on R2D and D2A tasks by 10%. |
Towards Fully Exploiting LLM Internal States to Enhance Knowledge Boundary Perception (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) exhibit impressive performance across diverse tasks but struggle to accurately gauge their knowledge boundaries. |
| Approach: | They propose Consistency-based Confidence Calibration (C3) which assesses confidence consistency through question reformulation to improve LLMs’ ability to recognize their knowledge gaps. |
| Outcome: | The proposed method improves the unknown perception rate by 5.6% on NQ and 4.9% on HotpotQA. |
ALGEN: Few-shot Inversion Attacks on Textual Embeddings via Cross-Model Alignment and Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have proven that private textual data is vulnerable to inversion attacks . authors present a method to reduce the cost of inversion and improve performance . |
| Approach: | They propose a method that aligns victim embeddings to attack space and reconstructs text . they find that none of the defense mechanisms are effective against inversion attacks . |
| Outcome: | The proposed method lowers the cost of inversion and improves performance across languages and domains. |
Decoding on Graphs: Faithful and Sound Reasoning on Knowledge Graphs through Generation of Well-Formed Chains (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research on the utilization of Knowledge Graphs (KGs) for large language models (LLMs) relies on subgraph retriever or iterative prompting, overlooking the potential synergy of LLMs’ step-wise reasoning capabilities and KGs’ structural nature. |
| Approach: | They propose a graph-aware constrained decoding framework that facilitates a deep synergy between LLMs and KGs by constraint derived from the topology of the KG. |
| Outcome: | The proposed framework can provide faithful and sound reasoning for KGQA. |
STaR-SQL: Self-Taught Reasoner for Text-to-SQL (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for generating step-by-step “chain-of-thought” rationales are limited to text-to-SQL. |
| Approach: | They propose a method that prompts SQL query generation to produce reasoning steps for SQL queries and fine-tunes it on rationales that lead to correct outcomes. |
| Outcome: | The proposed method outperforms agent-like prompting methods on the Spider benchmark. |
Fairness Beyond Performance: Revealing Reliability Disparities Across Groups in Legal NLP (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study shows that models often make less reliable or overconfident predictions for marginalized groups. |
| Approach: | They evaluate performance and reliability disparities across demographic, regional, and legal attributes across four jurisdictions using the FairLex benchmark. |
| Outcome: | The FairLex benchmark shows that pre-training improves performance and reliability for underrepresented groups. |
Beyond Similarity: A Gradient-based Graph Method for Instruction Tuning Data Selection (2025.acl-long)
Copied to clipboard
Yang Zhao, Li Du, Xiao Ding, Yangou Ouyang, Hepeng Wang, Kai Xiong, Jinglong Gao, Zhouhao Sun, Dongliang Xu, Qing Yang, Dongchen Li, Bing Qin, Ting Liu
| Challenge: | Existing methods for selecting training data from general datasets fail to account for the joint distribution of instructions, resulting in inefficient learning and suboptimal knowledge transfer. |
| Approach: | They propose a method that constructs a mixed gradient-based instruction graph to capture the joint distribution and interdependencies among instructions. |
| Outcome: | The proposed method outperforms existing methods on domain adaptation tasks and in complex, data-scarce scenarios. |
FastMCTS: A Simple Sampling Strategy for Data Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for generating multi-step reasoning data rely on rejection sampling, which generates trajectories independently and suffers from inefficiency and imbalanced sampling across problems of varying difficulty levels. |
| Approach: | They propose a data synthesis strategy inspired by Monte Carlo Tree Search . it offers step-level evaluation signals and promotes balanced sampling . |
| Outcome: | Experiments show that FastMCTS generates 30% more correct reasoning paths than rejection sampling. |
Dialogue-RAG: Enhancing Retrieval for LLMs via Node-Linking Utterance Rewriting (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) methods have demonstrated significant potential on tasks across multiple domains. |
| Approach: | They propose a lightweight IUR model for query rewriting to complete key information in dialogue to enhance retrieval. |
| Outcome: | The proposed model improves retrieval and generation ability of RAG system in multi-round dialogue scenarios. |
Using Information Theory to Characterize Prosodic Typology: The Case of Tone, Pitch-Accent and Stress-Accent (2025.acl-long)
Copied to clipboard
| Challenge: | lexical identity and prosody are well-studied parameters of linguistic variation, but they are difficult to predict in tonal languages. |
| Approach: | They propose to characterize the relationship between lexical identity and prosody using information theory to estimate mutual information between the text and pitch curves. |
| Outcome: | The proposed hypothesis supports perspectives that view linguistic typology as gradient, rather than categorical. |
Evaluating LLMs for Portuguese Sentence Simplification with Linguistic Insights (2025.acl-long)
Copied to clipboard
Arthur Mariano Rocha De Azevedo Scalercio, Elvis A. De Souza, Maria José Bocorny Finatto, Aline Paes
| Challenge: | Sentence simplification (SS) aims to make sentences more straightforward to read and understand without changing its key points. |
| Approach: | They compare 26 state-of-the-art LLMs in Portuguese SS with two simplification models trained explicitly for this task and language. |
| Outcome: | The proposed models outperform open-source models in Portuguese SS . the models are compared against two simplification models trained for Portuguese . |
LaTIM: Measuring Latent Token-to-Token Interactions in Mamba Models (2025.acl-long)
Copied to clipboard
| Challenge: | State space models lack interpretability tools for long-context sequence modeling. |
| Approach: | They propose a token-level decomposition method for Mamba-1 and Mamba-2 that enables fine-grained interpretability. |
| Outcome: | The proposed method is able to reveal Mamba’s token-to-token interaction patterns across multiple tasks including translation, copying, and retrieval-based generation. |
Improving Low-Resource Morphological Inflection via Self-Supervised Objectives (2025.acl-long)
Copied to clipboard
| Challenge: | Rapid progress in natural language processing (NLP) has largely been driven by training transformer models on massive amounts of unlabeled data, but such large datasets are scarce for many of the world's languages. |
| Approach: | They propose to train encoder-decoder transformers for 19 languages and 13 auxiliary objectives on massive amounts of unlabeled data. |
| Outcome: | The proposed tasks outperform standard CMLM in character-level tasks when available data is limited. |
Don’t Reinvent the Wheel: Efficient Instruction-Following Text Embedding based on Guided Space Transformation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for text embedding require re-encoding the entire corpus for each instruction. |
| Approach: | They propose a framework that generates dynamic text embeddings that adapt to user instructions, highlighting specific attributes of text. |
| Outcome: | The proposed framework improves instruction-following text embedding quality over state-of-the-art methods while speeding up processing on large datasets. |
BOOKCOREF: Coreference Resolution at Book Scale (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks for coreference resolution systems are limited in length and do not adequately assess system capabilities at the book scale. |
| Approach: | They propose a novel pipeline that produces high-quality coreference resolution annotations on full narrative texts and a book-scale benchmark, BOOKCOREF. |
| Outcome: | The proposed pipeline produces high-quality coreference resolution annotations on full texts with an average document length of more than 200,000 tokens. |
OMGM: Orchestrate Multiple Granularities and Modalities for Efficient Multimodal Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for Knowledge-Based Visual Question Answering lack multimodal retrieval . large language models (LLMs) have demonstrated remarkable generalization and reasoning capabilities in text-based systems. |
| Approach: | They propose a multimodal vision-language retrieval-augmented generation system that harmonizes multiple modalities and modality to enhance retrieval. |
| Outcome: | The proposed system achieves state-of-the-art retrieval performance and competitive answers on InfoSeek and Encyclopedic-VQA benchmarks. |
Alleviating Hallucinations from Knowledge Misalignment in Large Language Models via Selective Abstention Learning (2025.acl-long)
Copied to clipboard
Lei Huang, Xiaocheng Feng, Weitao Ma, Yuchun Fan, Xiachong Feng, Yuxuan Gu, Yangfan Ye, Liang Zhao, Weihong Zhong, Baoxin Wang, Dayong Wu, Guoping Hu, Lingpeng Kong, Tong Xiao, Ting Liu, Bing Qin
| Challenge: | Large language models (LLMs) suffer from severe hallucination issues due to the knowledge misalignment between the pre-training stage and the supervised fine-tuning stage. |
| Approach: | They propose a training objective with an abstention mechanism that selectively rejects tokens that misalign with the desired knowledge distribution via a special [REJ] token. |
| Outcome: | The proposed model selectively rejects tokens that misalign with the desired knowledge distribution via a special [REJ] token. |
Retrospective Learning from Interactions (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-turn interactions between large language models and users naturally include implicit feedback signals. |
| Approach: | They propose a method to learn from feedback signals in past interactions without annotations . they use a multimodal LLM to solve a reasoning task with a combinatorial solution space . |
| Outcome: | The proposed method improves task completion rate from 31% to 82% without annotations. |
Personalized Generation In Large Model Era: A Survey (2025.acl-long)
Copied to clipboard
Yiyan Xu, Jinghao Zhang, Alireza Salemi, Xinting Hu, Wenjie Wang, Fuli Feng, Hamed Zamani, Xiangnan He, Tat-Seng Chua
| Challenge: | Recent advances in large generative models have catalyzed a paradigm shift in content generation to Personalized Generation (PGen). |
| Approach: | They propose a multi-level taxonomy that systematically formalizes PGen's key components, core objectives, and abstract workflows. |
| Outcome: | The proposed taxonomy bridging PGen research across multiple modalities highlights open challenges and promising directions for future exploration. |
Graph Counselor: Adaptive Graph Exploration via Multi-Agent Synergy to Enhance LLM Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for enhancing LLM reliability suffer from inefficient information aggregation and rigid reasoning schemes. |
| Approach: | They propose a method that explicitly models external knowledge integration capabilities by explicitly modeling knowledge relationships. |
| Outcome: | The proposed method outperforms existing methods in multiple graph reasoning tasks. |
SOTOPIA-: Dynamic Strategy Injection Learning and Social Instruction Following Evaluation for Social Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on the social simulation of large language model intelligent agents have shown that even expert agents 1 perform significantly worse on challenging social tasks compared to expert agents. |
| Approach: | They propose a framework that dynamically injects a variety of social strategies into expert agents, thereby automating the construction of high-quality social dialogue training corpus. |
| Outcome: | The proposed framework enables the integration of social strategies into language agents and improves their performance on social tasks. |
Can Language Models Replace Programmers for Coding? REPOCOD Says ‘Not Yet’ (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks for code generation use short completions, synthetic examples, or focus on limited scale repositories, failing to represent real-world coding tasks. |
| Approach: | They propose a Python code-generation benchmark that contains 980 whole-function generation tasks with realistic dependencies from 11 popular projects. |
| Outcome: | The proposed benchmarks are short completions, synthetic examples, or focus on limited scale repositories, failing to represent real-world coding tasks. |
Leveraging In-Context Learning for Political Bias Testing of LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing probing methods for evaluating LLMs with political questions have limited stability and are unreliable. |
| Approach: | They propose to use human survey data as in-context examples to query LLMs with political questions to evaluate their potential biases. |
| Outcome: | The proposed task improves the stability of question-based bias evaluation and may be used to compare instruction-tuned models to their base versions. |
ACORD: An Expert-Annotated Retrieval Dataset for Legal Contract Drafting (2025.acl-long)
Copied to clipboard
Steven H Wang, Maksim Zubkov, Kexin Fan, Sarah Harrell, Yuyang Sun, Wei Chen, Andreas Plesner, Roger Wattenhofer
| Challenge: | Contract clause retrieval is critical to contract drafting because of its high quality and complexity. |
| Approach: | They propose the first expert-annotated benchmark specifically designed for contract clause retrieval . ACORD focuses on complex contract clauses such as Limitation of Liability, Indemnification, Change of Control . |
| Outcome: | The atticus clause retrieval dataset shows promising results but needs improvement . the benchmark can be used as an IR benchmark for the NLP community . |
LLMs know their vulnerabilities: Uncover Safety Gaps through Natural Distribution Shifts (2025.acl-long)
Copied to clipboard
Qibing Ren, Hao Li, Dongrui Liu, Zhanxu Xie, Xiaoya Lu, Yu Qiao, Lei Sha, Junchi Yan, Lizhuang Ma, Jing Shao
| Challenge: | Current safety training focuses on teaching models to reject harmful queries, but recent research shows that adversarial attacks or jailbreak methods bypass these safety mechanisms. |
| Approach: | They propose to use a new attack method to craft multi-turn toxic prompts that gradually lead LLMs to reveal unsafe content. |
| Outcome: | The proposed method outperforms existing methods in diversity, effectiveness, and efficiency across aligned LLMs. |
WAFFLE: Fine-tuning Multi-Modal Model for Automated Front-End Development (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown promise in generating source code, but two major challenges persist in UI-to-HTML code generation: (1) effectively representing HTML’s hierarchical structure for LLMs; and (2) bridging the gap between the visual nature of UI designs and the text-based format of HTML code. |
| Approach: | They propose a structure-aware attention mechanism that uses a contrastive fine-tuning approach to align LLMs’ understanding of UI images and HTML code. |
| Outcome: | The proposed model outperforms existing methods on the WebSight-Test and Design2Code benchmarks. |
Math Neurosurgery: Isolating Language Models’ Math Reasoning Abilities Using Only Forward Passes (2025.acl-long)
Copied to clipboard
| Challenge: | Math reasoning is an active area of Large Language Model (LLM) research because it is a hallmark of artificial intelligence and has implications in several domains, including math education. |
| Approach: | They propose a method to isolate math-specific parameters in LLMs using only forward passes. |
| Outcome: | The proposed method improves a model's performance on GSM8K and MATH by 4-17% while leaving non-math behavior unaltered. |
Multiple LLM Agents Debate for Equitable Cultural Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Recent efforts focus on single-LLM, single-turn generation approaches, but it can be challenging for any single model to support all cultures equally well. |
| Approach: | They propose to exploit the complementary strengths of multiple LLMs to promote cultural adaptability. |
| Outcome: | The proposed model improves accuracy and cultural group parity over single-LLM models. |
RefreshKV: Updating Small KV Cache During Long-form Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for generating long sequences of tokens are expensive and require memory and computation resources. |
| Approach: | They propose a method that alternates between full context attention and attention over a subset of input tokens during generation. |
| Outcome: | The proposed method achieves comparable speedup to eviction-based methods while improving performance for various long-form generation tasks. |
SEA: Low-Resource Safety Alignment for Multimodal Large Language Models via Synthetic Embeddings (2025.acl-long)
Copied to clipboard
| Challenge: | Existing low-resource security alignment methods struggle with the security risks posed by additional modalities. |
| Approach: | They propose to use multimodal datasets to enhance safety alignment but it is costly to construct these datasets. |
| Outcome: | Experiments on image, video, and audio-based MLLMs show that the proposed method can synthesize a high-quality embedding on a single RTX3090 GPU within 24 seconds. |
Chain-of-Reasoning: Towards Unified Mathematical Reasoning in Large Language Models via a Multi-Paradigm Perspective (2025.acl-long)
Copied to clipboard
Yiyao Yu, Yuxiang Zhang, Dongdong Zhang, Xiao Liang, Hengyuan Zhang, Xingxing Zhang, Mahmoud Khademi, Hany Hassan Awadalla, Junjie Wang, Yujiu Yang, Furu Wei
| Challenge: | Existing work shows that LLMs rely on single-paradigm reasoning that limits their effectiveness across diverse tasks. |
| Approach: | They propose a new framework that integrates multiple reasoning paradigms to enable synergistic collaboration. |
| Outcome: | The proposed model outperforms current SOTA models in theorem proving tasks and the MATH benchmark in arithmetic tasks. |
Language Models Grow Less Humanlike beyond Phase Transition (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have shown that LMs' alignment with human reading behavior improves during pretraining up to a tipping point, beyond which it plateaus or degrades. |
| Approach: | They hypothesize that a pretraining phase transition is responsible for the tipping point in PPP and that phase transitions alter the subsequent learning dynamics of the model, such that further training keeps damaging PPP. |
| Outcome: | The proposed model is able to produce attention patterns that contribute to the degradation of PPP, but it is not capable of producing attention patterns. |
PCoT: Persuasion-Augmented Chain of Thought for Detecting Fake News and Social Media Disinformation (2025.acl-long)
Copied to clipboard
| Challenge: | Psychological studies have shown that infusing persuasion knowledge enhances disinformation detection. |
| Approach: | They introduce a persuasion-augmented chain of thought approach that leverages persulasion to improve disinformation detection in zero-shot classification. |
| Outcome: | The proposed approach outperforms competitive methods by 15% on online news and social media posts. |
Coordinating Chaos: A Structured Review of Linguistic Coordination Methodologies (2025.acl-long)
Copied to clipboard
| Challenge: | Linguistic coordination is a phenomenon where conversation partners have similar patterns of language use. |
| Approach: | They propose a framework to organize the literature on linguistic coordination . they propose linguistic modeling choices and critiques of the choices involved . |
| Outcome: | The proposed framework provides an overview of the choices involved in the measurement process and synthesizes relevant critiques. |
iNews: A Multimodal Dataset for Modeling Personalized Affective Responses to News (2025.acl-long)
Copied to clipboard
| Challenge: | Current approaches to modeling individual behavior ignore individual differences in how people interpret and react to identical stimuli. |
| Approach: | They propose a large-scale dataset specifically designed to facilitate the modeling of personalized affective responses to news content. |
| Outcome: | The proposed dataset includes annotations from 291 demographically diverse UK participants across 2,899 multimodal Facebook news posts from major UK outlets. |
Mind the Gesture: Evaluating AI Sensitivity to Culturally Offensive Non-Verbal Gestures (2025.acl-long)
Copied to clipboard
| Challenge: | a dataset of 288 gesture-country pairs is used to evaluate AI systems' cultural awareness of offensive gestures and nonverbal signs. |
| Approach: | They use a dataset of 288 gesture-country pairs annotated for offensiveness, cultural significance, and contextual factors across 25 gestures and 85 countries. |
| Outcome: | The proposed dataset analyzes 288 gesture-country pairs across 25 gestures and 85 countries. |
500xCompressor: Generalized Prompt Compression for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Prompt compression is important for large language models to increase inference speed, reduce computation cost, and improve user experience. |
| Approach: | They propose a method that compresses natural language contexts into a special token . they propose to reduce computations and memory costs by reducing the complexity . |
| Outcome: | The proposed method reduces computations and memory costs by 27-90% . it retains 70-74% and 77-84% of the LLM capabilities at high compression ratios . |
Estimating Privacy Leakage of Augmented Contextual Knowledge in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Prior work has shown that privacy leakage of parametric knowledge often occurs from memorized pre-training data. |
| Approach: | They propose a metric that builds on differential privacy to estimate the privacy leakage of contextual knowledge during decoding by comparing parametric and contextual knowledge. |
| Outcome: | The proposed method overestimates the privacy leakage of parametric knowledge while separating parametric and contextual knowledge. |
Document-Level Event-Argument Data Augmentation for Challenging Role Types (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for Event Argument Extraction (EAE) are not well-suited to a variety of real-world situations, including long documents and challenging role types. |
| Approach: | They propose two novel methods for generating document-level EAE samples using zero in-domain training data and validate their generalizability. |
| Outcome: | The proposed methods show significant performance increases in low-resource settings. |
Mapping the Podcast Ecosystem with the Structured Podcast Research Corpus (2025.acl-long)
Copied to clipboard
| Challenge: | a dataset of over 1.1M podcast transcripts is largely comprehensive of all English language podcasts available through public RSS feeds from May and June of 2020. |
| Approach: | They propose to build a large-scale open dataset of podcast transcripts that includes metadata, speaker roles, audio features and speaker turns for a subset of 370K episodes. |
| Outcome: | The proposed dataset is largely comprehensive of all English language podcasts available through public RSS feeds from May and June of 2020. |
Unravelling the Logic: Investigating the Generalisation of Transformers in Numerical Satisfiability Problems (2025.acl-long)
Copied to clipboard
| Challenge: | Transformer models exhibit minimal scale and noise invariance, along with limited vocabulary and number invariancy. |
| Approach: | They probe the generalisation prowess of Transformer models with respect to the hitherto unexplored domain of numerical satisfiability problems. |
| Outcome: | The proposed models exhibit minimal scale and noise invariance, along with limited vocabulary and number invariancy. |
The Nature of NLP: Analyzing Contributions in NLP Papers (2025.acl-long)
Copied to clipboard
| Challenge: | despite this, what constitutes NLP research remains debated . |
| Approach: | They propose a taxonomy of research contributions and introduce a task of automatically identifying contribution statements and classifying their types from NLP research papers. |
| Outcome: | The proposed model analyzes 29k NLP research papers to understand their contributions . |
\mathtt{GeLLM^3O}: Generalizing Large Language Models for Multi-property Molecule Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have remarkable out-of-domain generalizability to novel optimization tasks. |
| Approach: | They propose a series of instruction-tuned LLMs for molecule optimization that outperform state-of-the-art instruction-based LLM models. |
| Outcome: | mathttMuMOInstruct outperforms state-of-the-art LLMs on 5 in-domain and 5 out-of domain tasks. |
Follow-up Question Generation For Enhanced Patient-Provider Conversations (2025.acl-long)
Copied to clipboard
Joseph Gatto, Parker Seegmiller, Timothy E. Burdick, Inas S. Khayal, Sarah DeLozier, Sarah Masud Preum
| Challenge: | Follow-up question generation is an essential feature of dialogue systems as it can reduce conversational ambiguity and enhance modeling complex interactions. |
| Approach: | They propose a framework that generates personalized follow-up questions based on patient utterances and prior EHR data. |
| Outcome: | The framework reduces follow-up communications by 34% and improves performance by 17% and 5% on real and synthetic data. |
Unveiling Privacy Risks in LLM Agent Memory (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Model (LLM) agents store private user-agent interactions in memory for demonstrations, introducing new privacy risks for LLM agents. |
| Approach: | They propose an attack that extracts private information from memory under a black-box setting and propose a method that can be used to attack the agent. |
| Outcome: | The proposed attack is effective under a black-box setting and it is demonstrated on two representative agents. |
Watching the Watchers: Exposing Gender Disparities in Machine Translation Quality Estimation (2025.acl-long)
Copied to clipboard
| Challenge: | Qualitative estimation (QE) metrics have been optimized to align with human quality judgments, but whether they encode social biases has been largely overlooked. |
| Approach: | They define and investigate gender bias of QE metrics and discuss its downstream implications for machine translation (MT) when a human entity’s gender in the source is undisclosed, masculine-inflected translations score higher than feminine-infflectes translations are penalized. |
| Outcome: | The proposed measures are based on gender-based quality estimation metrics across multiple domains, datasets, and languages. |
Language Constrained Multimodal Hyper Adapter For Many-to-Many Multimodal Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models that share parameters neglect the language-specific knowledge learning. |
| Approach: | They propose a language-constrained multimodal hyper adapter for multimodal summarization that integrates language-specific adapters into multilingual pre-trained backbones. |
| Outcome: | The proposed model can generate summaries based on multimodal documents such as text and visuals, allowing people to quickly locate key information from the vast multimedia con. |
PRMBench: A Fine-grained and Challenging Benchmark for Process-Level Reward Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent large language models (LLMs) have achieved significant performance in complex reasoning tasks such as mathematics and code generation. |
| Approach: | They propose a process-level benchmark specifically designed to assess the fine-grained error detection capabilities of PRMs. |
| Outcome: | The proposed model measures the accuracy, soundness, and sensitivity of 25 models across open-source and closed-source large language models. |
Efficient Ensemble for Fine-tuning Language Models on Multiple Datasets (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning language models are efficient when adapting to a single dataset. |
| Approach: | They propose to use an ensemble method for fine-tuning a language model to multiple datasets instead of a single adapter per task. |
| Outcome: | The proposed method improves performance on multiple datasets while preserving low-rank adaptation properties. |
Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles (2025.acl-long)
Copied to clipboard
Munachiso S Nwadike, Zangir Iklassov, Toluwani Aremu, Tatsuya Hiraoka, Benjamin Heinzerling, Velibor Bojkovic, Hilal AlQuabeh, Martin Takáč, Kentaro Inui
| Challenge: | Existing models that use sequential data can bypass the limitations of unidirectional causality. |
| Approach: | They propose a mechanism that enables large language models to bypass unidirectional causality . they propose 'cycle tokens' that enable recall of preceding tokens from succeeding ones . |
| Outcome: | The proposed model bypasses the limitations of unidirectional causality by enabling recall of preceding contexts. |
Shaping the Safety Boundaries: Understanding and Defending Against Jailbreaks in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Understanding how jailbreaking works remains limited, hindering the development of effective defense strategies. |
| Approach: | They propose a new mechanism that adaptively constrains activations within the safety boundary and propose 'Activation Boundary Defense' to enhance its effectiveness. |
| Outcome: | The proposed defense achieves an average Defense Success Rate (DSR) of over 98% against various jailbreak attacks, with less than 2% impact on the model’s general capabilities. |
ASPERA: A Simulated Environment to Evaluate Planning for Complex Action Execution (2025.acl-long)
Copied to clipboard
Alexandru Coca, Mark Gaynor, Zhenxing Zhang, Jianpeng Cheng, Bo-Hsiang Tseng, Peter Boothroyd, Hector Martinez Alonso, Diarmuid O Seaghdha, Anders Johannsen
| Challenge: | ASPERA framework allows developers to guide LLM generation of high-quality tasks based on user queries, simulation state and corresponding validation programs. |
| Approach: | They develop a framework comprising an assistant library simulation and a human-assisted LLM data generation engine to guide LLM generation of high-quality tasks . they use a dataset to evaluate 250 tasks generated using ASPERA . |
| Outcome: | The proposed framework can guide LLM generation of high-quality tasks tackling data availability and evaluation robustness challenges. |
ReflectDiffu: Reflect between Emotion-intent Contagion and Mimicry for Empathetic Response Generation via a RL-Diffusion Framework (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models for empathetic dialogue generation neglect the intricate interplay between emotion and intent, leading to suboptimal controllability of empathy. |
| Approach: | They propose a framework that integrates emotion contagion and intent mimicry to enhance empathetic response generation. |
| Outcome: | The proposed framework outperforms existing models in relevance, controllability, and informativeness. |
SARA: Salience-Aware Reinforced Adaptive Decoding for Large Language Models in Abstractive Summarization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing decoding strategies neglect the explicit use of salient contextual information and rely on static hyperparameters to fix the balance between contextual and prior knowledge. |
| Approach: | They propose a salience-aware reinforced adaptive decoding (SARA) which incorporates salient contextual information and allows the model to determine reliance on source document's context, salient context, and model's prior knowledge based on pointwise mutual information. |
| Outcome: | The proposed model improves the quality and faithfulness of summaries across LLMs without modifying their weights. |
Embedding-Converter: A Unified Framework for Cross-Model Embedding Transformation (2025.acl-long)
Copied to clipboard
| Challenge: | Embedding models are fundamental to modern machine learning, but the continuous development of new models presents a major challenge. |
| Approach: | They propose a framework for efficiently transforming embeddings between different models, thus avoiding costly ‘re-embedding’. |
| Outcome: | The proposed framework achieves 100 times faster and cheaper computations in real-world applications. |
Improving Automatic Evaluation of Large Language Models (LLMs) in Biomedical Relation Extraction via LLMs-as-the-Judge (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models generate human-like text, making them unreliable for biomedical relation extraction tasks. |
| Approach: | They propose to use Large Language Models as judges to evaluate biomedical relation extraction . they propose structured output formatting for LLM-generated responses that helps LLMs improve their performance by 15%. |
| Outcome: | The proposed method improves LLM-Judges' performance by 15% . it is cheaper and more efficient than human evaluation metrics, the authors say . |
Answering Complex Geographic Questions by Adaptive Reasoning with Visual Context and External Commonsense Knowledge (2025.acl-long)
Copied to clipboard
| Challenge: | a new task of answering geographic reasoning questions based on the given image is proposed . the task requires identifying the objects in the image and understanding the background context . |
| Approach: | They propose a task of answering geographic reasoning questions based on the given image . they analyze the image and describe its fine-grained content by text and keywords . |
| Outcome: | The proposed method can be used to answer geographic reasoning questions based on an image . it can be applied to a large-scale dataset with 41,329 samples . |
Safety Alignment via Constrained Knowledge Unlearning (2025.acl-long)
Copied to clipboard
Zesheng Shi, Yucheng Zhou, Jing Li, Yuxin Jin, Yu Li, Daojing He, Fangming Liu, Saleh Alharbi, Jun Yu, Min Zhang
| Challenge: | Existing defense mechanisms have not fully deleted harmful knowledge in large language models (LLMs) Existing methods to address safety alignment have not completely deleted harmful information in LLMs. |
| Approach: | They propose a safety alignment strategy that uses scoring neurons to identify useful knowledge in LLMs and pruning the gradients of neurons in U to preserve beneficial information. |
| Outcome: | The proposed method significantly improves model safety while maintaining utility compared to existing methods. |
Response Wide Shut? Surprising Observations in Basic Vision Language Model Capabilities (2025.acl-long)
Copied to clipboard
| Challenge: | Vision-language Models have been shown to be highly capable but lacking basic visual understanding skills. |
| Approach: | They propose to examine the limitations of vision-language models on visual tasks by constructing a series of tests that probe which components of design may be lacking. |
| Outcome: | The proposed tests compare VLMs to other models on visual encoders, intermediate vision-language projection and LLM-decoder outputs. |
EffiVLM-BENCH: A Comprehensive Benchmark for Evaluating Training-Free Acceleration in Large Vision-Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for accelerating Large Vision-Language Models lack comprehensive evaluation across diverse backbones, benchmarks, and metrics. |
| Approach: | They propose EffiVLM-BENCH framework for evaluating absolute performance and generalization and loyalty. |
| Outcome: | The proposed framework offers insights into optimal strategies for accelerating LVLMs. |
Pre-Training Curriculum for Multi-Token Prediction in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-token prediction (MTP) is a pre-training objective for language models . prior work has shown that smaller language models struggle with the MTP objective . |
| Approach: | They propose a curriculum learning strategy that uses multiple prediction heads to predict the next tokens at each prediction step. |
| Outcome: | The proposed curriculum improves performance and output quality while retaining the benefits of self-speculative decoding. |
Can We Further Elicit Reasoning in LLMs? Critic-Guided Planning with Retrieval-Augmentation for Solving Challenging Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to problem-solving for large language models fail to provide accurate reasoning and factual accuracy. |
| Approach: | They propose a framework that leverages fine-tuned critic models to guide reasoning and retrieval processes. |
| Outcome: | The proposed framework outperforms baselines on domain-knowledge-intensive tasks . it can be used to iterate retrieval and reasoning, and improve retrieval relevance . |
On Many-Shot In-Context Learning for Long-Context Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks primarily evaluate long-context language models' retrieval capabilities. |
| Approach: | They propose a benchmark to evaluate long-context language models' retrieval capabilities by using MANYICLBENCH. |
| Outcome: | The proposed model performs better with additional demonstrations than translation and reasoning tasks. |
HelpSteer3: Human-Annotated Feedback and Edit Data to Empower Inference-Time Scaling in Open-Ended General-Domain Tasks (2025.acl-long)
Copied to clipboard
Zhilin Wang, Jiaqi Zeng, Olivier Delalleau, Daniel Egert, Ellie Evans, Hoo-Chang Shin, Felipe Soares, Yi Dong, Oleksii Kuchaiev
| Challenge: | Inference-Time Scaling is critical to the success of recent models such as OpenAI o1 and DeepSeek R1 . however, many techniques require tasks to have answers that can be verified . |
| Approach: | They use data to train dedicated Feedback and Edit Models capable of inference-time scaling for open-ended tasks. |
| Outcome: | The proposed model can reach SoTA performance on Arena Hard at 92.7 as of 5 Mar 2025. |
CulturalBench: A Robust, Diverse and Challenging Benchmark for Measuring LMs’ Cultural Knowledge Through Human-AI Red-Teaming (2025.acl-long)
Copied to clipboard
Yu Ying Chiu, Liwei Jiang, Bill Yuchen Lin, Chan Young Park, Shuyue Stella Li, Sahithya Ravi, Mehar Bhatia, Maria Antoniak, Yulia Tsvetkov, Vered Shwartz, Yejin Choi
| Challenge: | CulturalBench is a set of 1,696 human-written and human-verified questions to assess LMs’ cultural knowledge covering 45 global regions including underrepresented ones like Bangladesh, Zimbabwe, and Peru. |
| Approach: | They construct a set of 1,696 human-written and human-verified questions to assess LMs' cultural knowledge, covering 45 global regions including underrepresented ones like Bangladesh, Zimbabwe, and Peru. |
| Outcome: | The proposed model outperforms other models across cultures, while underperforming on questions related to North Africa, South America and Middle East. |
Balancing the Budget: Understanding Trade-offs Between Supervised and Preference-Based Finetuning (2025.acl-long)
Copied to clipboard
| Challenge: | Results show that supervised fine-tuning and preference finetunation are the most efficient approaches for large language models. |
| Approach: | They propose to use Supervised Finetuning and Preference Finetunes to optimize training data budgets for Large Language Models. |
| Outcome: | The proposed approach improves performance on math tasks by 15% on the most expensive model, 1,000 examples. |
All That Glitters is Not Novel: Plagiarism in AI Generated Research (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies claim autonomous research agents can generate novel research ideas. |
| Approach: | They ask experts to evaluate whether existing work is similar to new ones . they find 24% of the 50 evaluated documents to be either paraphrased or significantly borrowed . |
| Outcome: | The authors find that 24% of the 50 evaluated research documents are either paraphrased, or significantly borrowed from existing work. |
Writing Like the Best: Exemplar-Based Expository Text Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for expository text generation rely on extensive domain corpora and are difficult to adapt topic-specific content. |
| Approach: | They propose a task to generate an expository text on a new topic using an exemplar on . they propose Adaptive Imitation and a Recurrent Plan-then-Adapt framework that leverages large language models for effective adaptive imitation. |
| Outcome: | The proposed framework surpasses existing baselines in producing factual, consistent, and relevant texts for this task. |
Temporal Relation Extraction in Clinical Texts: A Span-based Graph Transformer Approach (2025.acl-long)
Copied to clipboard
| Challenge: | Temporal information extraction from unstructured text is challenging due to complex clinical language, long documents, and sparse annotations. |
| Approach: | They propose a method for extracting clinical events and their temporal relations from unstructured text using the I2B2 2012 Temporal Relations Challenge corpus. |
| Outcome: | The proposed method improves state-of-the-art temporal information extraction with 5.5% improvement in tempeval F1 score over previous best and 8.9% improvement on long-range relations. |
Finding A Voice: Exploring the Potential of African American Dialect and Voice Generation for Chatbots (2025.acl-long)
Copied to clipboard
| Challenge: | This study examines how linguistic similarity affects chatbot performance, focusing on integrating African American English (AAE) into virtual agents to better serve the African American community. |
| Approach: | They develop text-based and spoken chatbots using large language models and text-to-speech technology and evaluate them with AAE speakers to better serve the African American community. |
| Outcome: | The proposed language-based chatbots with African American English speakers outperform standard English chatbot models and show that spoken chatbot features improve performance and preference. |
Delta-KNN: Improving Demonstration Selection in In-Context Learning for Alzheimer’s Disease Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for in-context learning (ICL) perform poorly for AD diagnosis due to inherent complexity of task. |
| Approach: | They propose a demonstration selection strategy that leverages a delta score to assess the relative gains of each training example and a KNN-based retriever that dynamically selects optimal “representatives” for a given input. |
| Outcome: | The proposed model outperforms existing methods on two AD detection datasets and surpasses even supervised classifiers. |
Help Me Write a Story: Evaluating LLMs’ Ability to Generate Writing Feedback (2025.acl-long)
Copied to clipboard
| Challenge: | Current models provide specific and mostly accurate writing feedback, but they fail to identify the biggest writing issue in the story and to correctly decide when to offer critical vs. positive feedback. |
| Approach: | They propose a task that corrupts 1,300 stories to intentionally introduce writing issues to study model performance. |
| Outcome: | The proposed model performs well in a controlled task with human and automatic evaluation metrics. |
Language Fusion for Parameter-Efficient Cross-lingual Transfer (2025.acl-long)
Copied to clipboard
| Challenge: | Limited availability of multilingual text corpora for pretraining results in poor performance on downstream tasks due to undertrained representation spaces for languages other than English. |
| Approach: | They propose a method that integrates source and target language representations within low-rank (LoRA) adapters using lightweight linear transformations to enhance representation quality and transfer performance for languages other than English. |
| Outcome: | The proposed method improves representation quality and performance for languages other than English while maintaining parameter efficiency. |
Culture is Not Trivia: Sociocultural Theory for Cultural NLP (2025.acl-long)
Copied to clipboard
| Challenge: | Cultural NLP has experienced rapid growth to meet the need to ensure language technologies are effective and safe across a pluralistic user base. |
| Approach: | They propose to use a well-developed theory of culture to clarify methodological constraints and affordances and offer theoretically-motivated paths forward to achieving cultural competence. |
| Outcome: | The proposed framework clarifies methodological constraints and affordances and offers theoretically-motivated paths forward to achieving cultural competence. |
AAD-LLM: Neural Attention-Driven Auditory Scene Understanding (2025.acl-long)
Copied to clipboard
Xilin Jiang, Sukru Samet Dindar, Vishal Choudhari, Stephan Bickel, Ashesh Mehta, Guy M McKhann, Daniel Friedman, Adeen Flinker, Nima Mesgarani
| Challenge: | Existing auditory foundation models do not incorporate this selective focus on specific speakers . existing models do this by focusing on one speaker, ignoring others in complex auditory scenes . |
| Approach: | They present a prototype auditory scene understanding system that integrates brain signals to infer listener attention. |
| Outcome: | The proposed system integrates brain signals to infer listener attention . it evaluates speaker description, speech transcription and extraction, and question answering in multitalker scenarios . |
Do Language Models Have Semantics? On the Five Standard Positions (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are trained to solve the so-called cloze task . solving clozing tasks is essentially a memorization task, says a recent study . |
| Approach: | They propose to use five positions to determine whether large language models exhibit semantic understanding . large language model is trained to solve the so-called cloze task . |
| Outcome: | The proposed theory is based on a pairwise comparison of five positions on semantic understanding in large language models and chatbots. |
Dehumanizing Machines: Mitigating Anthropomorphic Behaviors in Text Generation Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have focused on how text generation systems can lead to harmful outcomes such as over-reliance, emotional dependence, dehumanization, deception, or even physical harm. |
| Approach: | They propose to use an inventory of interventions to help identify possible interventions and provide a conceptual framework to help characterize the landscape of possible interventions. |
| Outcome: | The proposed interventions are based on an inventory of interventions grounded in prior literature and a crowdsourcing study where participants edited system outputs to make them less human-like. |
Evaluating Multimodal Language Models as Visual Assistants for Visually Impaired Users (2025.acl-long)
Copied to clipboard
Antonia Karamolegkou, Malvina Nikandrou, Georgios Pantazopoulos, Danae Sanchez Villegas, Phillip Rust, Ruchira Dhar, Daniel Hershcovich, Anders Søgaard
| Challenge: | Despite high adoption rate of Large Language Models, there are limitations related to contextual understanding, cultural sensitivity, and complex scene understanding. |
| Approach: | They conduct a user survey to identify adoption patterns and key challenges users face with such technologies. |
| Outcome: | The proposed models have high adoption rates but still face limitations in visual aids. |
HumT DumT: Measuring and controlling human-like language in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Human-like language might improve user experience, but might also lead to deception, overreliance, and stereotyping. |
| Approach: | They introduce HumT and SocioT, metrics for human-like tone in LLM outputs . HumT measures human-type tone and other dimensions of social perceptions in text data . |
| Outcome: | The proposed method reduces human-like tone while preserving model performance. |
ChatBench: From Static Benchmarks to Human-AI Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | In 2024, 40% of US adults reported using generative AI in their everyday lives, an unprecedented rate of adoption for a new technology. |
| Approach: | They propose to convert MMLU questions into user-AI conversations by seeding the user with the question and having them carry out a conversation with the LLM to answer their question. |
| Outcome: | The proposed model can estimate user-AI accuracy by fine-tuning a user simulator on a subset of ChatBench. |
Teaching an Old LLM Secure Coding: Localized Preference Optimization on Distilled Preferences (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve security of LLM generated code are ineffective and lack localized regions of code. |
| Approach: | They propose a method for distilling a preference dataset of insecure and secure code pairs from frontier LLMs and a security reasoning that explains the issues and the fix. |
| Outcome: | The proposed method reduces code insecurity while improving overall code quality. |
Anything Goes? A Crosslinguistic Study of (Im)possible Language Learning in LMs (2025.acl-long)
Copied to clipboard
| Challenge: | LMs are highly flexible learners, capable of acquiring linguistic patterns beyond those learnable by humans. |
| Approach: | They train LMs to model impossible and typologically unattested languages . they find that the model does not achieve perfect separation between attested and unattest languages - suggesting some human-like inductive biases . |
| Outcome: | The proposed model can largely distinguish attested from impossible languages, but does not achieve perfect separation between them and their impossible counterparts. |
Ranking Unraveled: Recipes for LLM Rankings in Head-to-Head AI Combat (2025.acl-long)
Copied to clipboard
| Challenge: | Evaluating large language models (LLMs) is a complex task. Pairwise ranking has emerged as state-of-the-art method to evaluate human preferences. |
| Approach: | They propose to use pairwise ranking to evaluate human preferences . they propose to evaluate the robustness of ranking algorithms in LLMs . |
| Outcome: | The proposed methods are based on the principles of effective ranking and the robustness of several ranking algorithms in the context of LLMs. |
LLM Agents Making Agent Tools (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) can perform multi-step tasks by dynamically utilising external software components. |
| Approach: | They propose an agentic framework that autonomously transforms papers with code into LLM-compatible tools. |
| Outcome: | The proposed framework outperforms current state-of-the-art software engineering agents in 80% of tasks and is openly available on GitHub. |
CrafText Benchmark: Advancing Instruction Following in Complex Multimodal Open-Ended World (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to assess instruction following in dynamic and uncertain environments are limited and limited in their ability to adapt to the world's volatility and interdependencies. |
| Approach: | They propose a benchmark for evaluating instruction following in a multimodal environment with diverse instructions and dynamic interactions. |
| Outcome: | The proposed method measures an agent’s ability to generalize to novel instruction formulations and dynamically evolving task configurations, providing a rigorous test of both linguistic understanding and adaptive decision-making. |
QG-SMS: Enhancing Test Item Analysis via Student Modeling and Simulation (2025.acl-long)
Copied to clipboard
| Challenge: | Question Generation (QG) tasks are often evaluated using reference-based metrics such as ROUGE and BLEU. |
| Approach: | They propose a QG evaluation framework that leverages Large Language Model for Student Modeling and Simulation to perform test item analysis. |
| Outcome: | The proposed framework improves the QG task and human-simulated student profiles. |
Causal Graph based Event Reasoning using Semantic Relation Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in event reasoning have limited ability to accurately identify causal connections between events. |
| Approach: | They propose a collaborative approach to generate correct graphs and graphs to assist reasoning . they propose 'a causal chain of events' task that requires a causal link between events . |
| Outcome: | The proposed approach achieves competitive results with state-of-the-art models on forecasting and next event prediction tasks. |
LogicPro: Improving Complex Logical Reasoning via Program-Guided Learning (2025.acl-long)
Copied to clipboard
Jin Jiang, Yuchen Yan, Yang Liu, Jianing Wang, Shuai Peng, Xunliang Cai, Yixin Cao, Mengdi Zhang, Liangcai Gao
| Challenge: | LogicPro is a data synthesis method that uses LeetCode-style algorithm problems and their corresponding Program solutions to generate complex logic data. |
| Approach: | They propose a new method which leverages LeetCode-style algorithm Problems and their corresponding Program solutions to synthesize complex logic data in text format. |
| Outcome: | The proposed method outperforms existing models for BBH27, LogicBench, DROP, AR-LSAT, and GSM8K, and a wide range of reasoning datasets. |
Do LLMs Understand Dialogues? A Case Study on Dialogue Acts (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown remarkable performance on many unseen tasks in a zero-shot setting. |
| Approach: | They propose to identify three key pre-tasks essential for accurate DA prediction: Turn Management, Communicative Function Identification, and Dialogue Structure Prediction. |
| Outcome: | The proposed model fails to outperform basic rule-based tasks on three key pre-tasks, and the results suggest that the model is flawed. |
Research Borderlands: Analysing Writing Across Research Cultures (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study has focused on improving cultural competence of language technologies, but most studies rely on synthetic setups and imperfect proxies of culture. |
| Approach: | They use a human-centered approach to discover and measure language-based cultural norms and cultural competence of large language models (LLMs). |
| Outcome: | The proposed framework identifies cultural norms that vary across research cultures and identifie a lack of cultural competence in LLMs. |
CEAES: Bidirectional Reinforcement Learning Optimization for Consistent and Explainable Essay Assessment (2025.acl-long)
Copied to clipboard
| Challenge: | Current automated essay quality assessment systems treat score prediction and feedback generation as separate tasks. |
| Approach: | They propose a bidirectional reinforcement learning framework that jointly optimizes score prediction and feedback generation. |
| Outcome: | The proposed framework outperforms current state-of-the-art models in both scoring and feedback quality. |
DeAL: Decoding-time Alignment for Large Language Models (2025.acl-long)
Copied to clipboard
James Y. Huang, Sailik Sengupta, Daniele Bonadiman, Yi-An Lai, Arshit Gupta, Nikolaos Pappas, Saab Mansour, Katrin Kirchhoff, Dan Roth
| Challenge: | Large Language Models (LLMs) are expected to generate content aligned with human preferences. |
| Approach: | They propose a framework that allows the user to customize reward functions and enables Decoding-time Alignment of LLMs (DeAL). |
| Outcome: | The proposed framework allows the user to customize reward functions and enables Decoding-time Alignment of LLMs. |
Cultural Bias Matters: A Cross-Cultural Benchmark Dataset and Sentiment-Enriched Model for Understanding Multimodal Metaphors (2025.acl-long)
Copied to clipboard
| Challenge: | Metaphors are pervasive in communication, making them crucial for natural language processing. |
| Approach: | They propose a multicultural multimodal metaphor dataset designed for cross-cultural studies of metaphor in Chinese and English. |
| Outcome: | The proposed model improves metaphor comprehension across cultural backgrounds and cultural domains. |
OmniCharacter: Towards Immersive Role-Playing Agents with Seamless Speech-Language Personality Interaction (2025.acl-long)
Copied to clipboard
Haonan Zhang, Run Luo, Xiong Liu, Yuchuan Wu, Ting-En Lin, Pengpeng Zeng, Qiang Qu, Feiteng Fang, Min Yang, Lianli Gao, Jingkuan Song, Fei Huang, Yongbin Li
| Challenge: | Existing methods focus on replicating dialogues in textual form, neglecting the role’s voice traits as a crucial effect in interaction, which tends to be more immersive experiences in realistic scenarios. |
| Approach: | They propose a first seamless speech-language personality interaction model to achieve immersive RPAs with low latency. |
| Outcome: | The proposed model exhibits role-specific personality traits and vocal traits throughout the interaction, enabling a mixture of speech and language responses. |
Mixtures of In-Context Learners (2025.acl-long)
Copied to clipboard
| Challenge: | In-context learning is sensitive to the choice of in-con context demonstrations and processing many demonstrations can be computationally demanding. |
| Approach: | They propose a method that uses subsets of demonstrations to train experts via ICL and learns a weighting function to merge their output distributions via gradient-based optimisation. |
| Outcome: | The proposed approach improves on 5 out of 7 classification datasets compared to strong baselines and reduces the inference time needed to achieve the same performance with fewer demonstrations. |
Balancing Diversity and Risk in LLM Sampling: How to Select Your Method and Parameter for Open-Ended Text Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies suggest that sampling-based decoding strategies can be used to optimize the output of Large Language Models (LLMs) . previous studies have shown that likelihood-maximization produces degenerate text which contains repetitive loops and incoherent context, especially in open-ended tasks. |
| Approach: | They propose to use a prefix tree to estimate the intrinsic capacity of a truncation sampling method by considering the trade-off between diversity and risk at each decoding step. |
| Outcome: | The proposed method is based on a prefix tree which preserves the context of a full sentence. |
RADAR: Enhancing Radiology Report Generation with Supplementary Knowledge Injection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to enhance radiology report generation overlook the knowledge already embedded within the models, leading to redundant information integration. |
| Approach: | They propose a framework for enhancing radiology report generation with supplementary knowledge injection that leverages both internal and external knowledge. |
| Outcome: | Extensive experiments on MIMIC-CXR, CheXpert-Plus, and IU X-ray show that the proposed model outperforms state-of-the-art LLMs in both language quality and clinical accuracy. |
Can LLMs Deceive CLIP? Benchmarking Adversarial Compositionality of Pre-trained Multimodal Representation via Text Updates (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in multimodal systems have demonstrated remarkable capabilities in generating multimodal content from multimodal inputs. |
| Approach: | They propose a benchmark that leverages large language models to generate deceptive text samples to exploit compositional vulnerabilities across different modalities. |
| Outcome: | The proposed approach exploits compositional vulnerabilities across images, videos, and audios. |
Attention Speaks Volumes: Localizing and Mitigating Bias in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | ambiguous comparative prompting frameworks generate biased content when confronted with ambiguously prompts . ambiguity in LLMs can manifest in various forms, including reinforcing stereotypes, displaying gender bias, or demonstrating preferential treatment towards specific demographic groups. |
| Approach: | They propose a technique to localize bias to specific layers of an LLM by analyzing attention scores and then reduce bias by scaling attention in these biased layers. |
| Outcome: | The proposed technique reduces bias by scaling attention in biased layers. |
MTSA: Multi-turn Safety Alignment for LLMs through Multi-round Red-teaming (2025.acl-long)
Copied to clipboard
| Challenge: | Existing jailbreak techniques rely on single-round interactions, pro-Corresponding author. |
| Approach: | They propose a multi-turn safety alignment framework to address the challenge of securing large language models in multi-round interactions. |
| Outcome: | The proposed framework exhibits state-of-the-art attack capabilities while improving safety performance on safety benchmarks. |
The Efficiency vs. Accuracy Trade-off: Optimizing RAG-Enhanced LLM Recommender Systems Using Multi-Head Early Exit (2025.acl-long)
Copied to clipboard
Huixue Zhou, Hengrui Gu, Zaifu Zhan, Xi Liu, Kaixiong Zhou, Yongkang Xiao, Mingfu Liang, Srinivas Prasad Govindan, Piyush Chawla, Jiyan Yang, Xiangfei Meng, Huayu Li, Buyun Zhang, Liang Luo, Wen-Yen Chen, Yiping Han, Bo Long, Rui Zhang, Tianlong Chen
| Challenge: | Existing frameworks for Large Language Models (LLMs) for Click-Through Rate prediction require a careful balance between computational efficiency and predictive accuracy. |
| Approach: | They propose a framework that integrates Retrieval-Augmented Generation with a novel multi-head early exit architecture to address both challenges. |
| Outcome: | The proposed framework reduces retrieval time while maintaining high model performance. |
Unraveling LoRA Interference: Orthogonal Subspaces for Robust Model Merging (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning large language models fail due to performance degradation . existing methods fail for models fine- tuned with low-rank adaptation . |
| Approach: | They propose to constrain the LoRA subspace prior to fine-tuning to ensure that updates relevant to one task do not adversely shift outputs for others. |
| Outcome: | The proposed method can integrate with most existing merging algorithms, reducing unintended interference among tasks. |
BIG-Bench Extra Hard (2025.acl-long)
Copied to clipboard
Mehran Kazemi, Bahare Fatemi, Hritik Bansal, John Palowitch, Chrysovalantis Anastasiou, Sanket Vaibhav Mehta, Lalit K Jain, Virginia Aglietti, Disha Jindal, Peter Chen, Nishanth Dikkala, Gladys Tyen, Xin Liu, Uri Shalit, Silvia Chiappa, Kate Olszewska, Yi Tay, Vinh Q. Tran, Quoc V Le, Orhan Firat
| Challenge: | Current benchmarks for large language model reasoning focus on math and coding abilities, leaving a gap in evaluating broader reasoning proficiencies. |
| Approach: | They propose a benchmark to evaluate general reasoning in large language models . they use BIG-Bench and its harder version BIG-Benefit Hard to assess general reasoning . |
| Outcome: | The new benchmark pushes the boundaries of LLM reasoning evaluation. |
CSTree-SRI: Introspection-Driven Cognitive Semantic Tree for Multi-Turn Question Answering over Extra-Long Contexts (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have achieved remarkable success in natural language processing (NLP), particularly in single-turn question answering (QA) on short-text. |
| Approach: | They propose a framework that captures logical correlations across chunks of ELC and maintains coherence of multi-turn Questions. |
| Outcome: | The proposed framework is able to capture logical correlations across chunks of ELC and maintain coherence of multi-turn Questions. |
InductionBench: LLMs Fail in the Simplest Complexity Class (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks focus on deductive reasoning, largely overlooking inductive reasoning. |
| Approach: | They propose a benchmark to evaluate the inductive reasoning ability of large language models. |
| Outcome: | The proposed benchmark demonstrates that even the most advanced modelw struggle to master the simplest complexity classes within the subregular hierarchy of functions. |
RATIONALYST: Pre-training Process-Supervision for Improving Reasoning (2025.acl-long)
Copied to clipboard
Dongwei Jiang, Guoxuan Wang, Yining Lu, Andrew Wang, Jingyu Zhang, Chuyu Liu, Benjamin Van Durme, Daniel Khashabi
| Challenge: | RATIONALYST is a model for process-supervision of reasoning based on pretraining on rationale annotations extracted from unlabeled data. |
| Approach: | They propose a model for process-supervision of reasoning based on pre-training on rationale annotations extracted from unlabeled data. |
| Outcome: | RATIONALYST improves reasoning accuracy by 3.9% on representative reasoning benchmarks. |
Make Imagination Clearer! Stable Diffusion-based Visual Imagination for Multimodal Machine Translation (2025.acl-long)
Copied to clipboard
Andong Chen, Yuchen Song, Kehai Chen, Xuefeng Bai, Muyun Yang, Liqiang Nie, Jie Liu, Tiejun Zhao, Min Zhang
| Challenge: | Experimental results show that our model significantly outperforms existing multimodal MT and text-only MT. |
| Approach: | They propose a stable diffusion-based imagination network into a multimodal large language model to generate an image for each source sentence. |
| Outcome: | The proposed model outperforms existing multimodal and text-only MT and achieves an average improvement of 14 BLEU points on Multi30K and MSCOCO multimodal MT benchmarks. |
Advancing SMoE for Continuous Domain Adaptation of MLLMs: Adaptive Router and Domain-Specific Loss (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have explored Continual Instruction Tuning (CIT) in Multimodal Large Language Models (MLLMs), with a primary focus on Task-incremental CIT, where MLLM are required to continuously acquire new tasks. |
| Approach: | They propose a Sparse Mixture of Expert (SMoE) based method for domain-incremental CIT in Multimodal Large Language Models (MLLMs) . they equip the SMoA module with a domain-specific autoregressive loss (DSAL) they establish a new benchmark to evaluate the efficacy of their method . |
| Outcome: | The proposed method outperforms all baselines and is based on a Sparse Mixture of Experts (SMoE) module . |
Multi-document Summarization through Multi-document Event Relation Graph Reasoning in LLMs: a case study in Framing Bias Mitigation (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study has focused on detecting media bias in news articles . a multi-document event relation graph is used to generate a neutralized summary . |
| Approach: | They propose to generate a neutralized summary given multiple articles presenting different ideological views. |
| Outcome: | The proposed method mitigates media bias and improves content preservation. |
Who Writes What: Unveiling the Impact of Author Roles on AI-generated Text Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) require accurate text detection, but authors' characteristics are neglected. |
| Approach: | They investigate how author characteristics impact AI-generated text detection . they use corpus of human-authored texts and parallel AI-generated texts . |
| Outcome: | The results show that gender, CEFR proficiency, academic field and language environment influence detector accuracy. |
RoCoFT: Efficient Finetuning of Large Language Models with Row-Column Updates (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning large language models use full finetunation, but this is impractical as language models continue to scale up. |
| Approach: | They propose a parameter-efficient fine-tuning method for large language models based on updating only a few rows and columns of the weight matrices in transformers. |
| Outcome: | The proposed method gives comparable or better accuracies than state-of-the-art methods while being more memory and computation-efficient. |
Scaling Laws and Efficient Inference for Ternary Language Models (2025.acl-long)
Copied to clipboard
Tejas Vaidhya, Ayush Kaushal, Vineet Jain, Francis Couture-Harpin, Prashant Shishodia, Majid Behbahani, Yuriy Nevmyvaka, Irina Rish
| Challenge: | Large language models (LLMs) are increasingly used across research and industry applications, yet their inference efficiency remains a challenge. |
| Approach: | They propose ternary language models that employ quantization-aware training to significantly reduce memory requirements. |
| Outcome: | The proposed ternary language models demonstrate sustained performance gains at scale. |
Exploring the Impact of Instruction-Tuning on LLM’s Susceptibility to Misinformation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies highlight that large language models are receptive to external information that contradicts their parametric knowledge, but little research has been conducted on the direct impact of instruction-tuning on this phenomenon. |
| Approach: | They examine how instruction-tuning influences LLMs' susceptibility to misinformation, particularly in knowledge conflict situations. |
| Outcome: | The proposed model is more user-oriented and more likely to accept misinformation when it is presented by the user. |
Do Language Models Understand Honorific Systems in Javanese? (2025.acl-long)
Copied to clipboard
Mohammad Rifqi Farhansyah, Iwan Darmawan, Adryan Kusumawardhana, Genta Indra Winata, Alham Fikri Aji, Derry Tanti Wijaya
| Challenge: | Despite its cultural and linguistic significance, there has been limited progress in developing a comprehensive corpus to capture these variations for natural language processing (NLP) tasks. |
| Approach: | They propose to use a dataset to capture the nuances of Unggah-Ungguh Basa, the Javanese speech etiquette framework, to assess the ability of language models to process various levels of Javanesi honorifics. |
| Outcome: | The proposed dataset encapsulates the nuances of Unggah-Ungguh Basa, the Javanese speech etiquette framework. |
Generative Reward Modeling via Synthetic Criteria Preference Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Generative Reward Models (GenRMs) leverage synthesized Chains of Thought (CoT) but this approach introduces risks of overoptimization due to the inability to guarantee the correctness of the CoTs. |
| Approach: | They propose a criteria-based preference tree for GenRMs that uses chain of thought to generate reasoning . they show that synthesized data can be learned using a long CoT format . |
| Outcome: | The proposed model shows significant improvements over baselines on multiple human preference benchmarks. |
Exploring Multimodal Relation Extraction of Hierarchical Tabular Data with Multi-task Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies overlook the need of mining relations among multiple columns rather than just the semantic relation between two specific columns in real-world practice. |
| Approach: | They propose a Chain-of-Thought distillation framework with self-correction mechanism to enhance MLLMs’ reasoning capabilities without increasing parameter scale. |
| Outcome: | The proposed method significantly outperforms baselines on wide datasets. |
A Self-Denoising Model for Robust Few-Shot Relation Extraction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies assume that the support set contains only accurately labeled instances, but this assumption is often unrealistic. |
| Approach: | They propose a self-denoising model for FSRE which can automatically correct noisy labels of support instances. |
| Outcome: | The proposed model outperforms all baselines on two public datasets showing that it can correct mislabeled support instances. |
QuASAR: A Question-Driven Structure-Aware Approach for Table-to-Text Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for table-to-text generation fail to capture the structure of tabular data or rely on complex attention mechanisms, limiting their applicability. |
| Approach: | They propose a question-driven self-supervised approach to enhance the model’s structural perception and representation capabilities by focusing on structure-related queries. |
| Outcome: | The proposed model improves its model's structural perception and representation capabilities by guiding it to capture local and global table structures. |
Automated Structured Radiology Report Generation (2025.acl-long)
Copied to clipboard
Jean-Benoit Delbrouck, Justin Xu, Johannes Moll, Alois Thomas, Zhihong Chen, Sophie Ostmeier, Asfandyar Azhar, Kelvin Zhenghao Li, Andrew Johnston, Christian Bluethgen, Eduardo Pontes Reis, Mohamed S Muneer, Maya Varma, Curtis Langlotz
| Challenge: | Existing models struggle to produce consistent, clinically meaningful reports and standard evaluation metrics fail to capture the nuances of radiological interpretation. |
| Approach: | They propose to reformulate free-text radiology reports into a standardized format, ensuring clarity, consistency, and structured clinical reporting. |
| Outcome: | The proposed task reformulates free-text radiology reports into a standardized format, ensuring clarity, consistency, and structured clinical reporting. |
LPOI: Listwise Preference Optimization for Vision Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for aligning large VLMs with human preferences often overfit to textual information or exacerbate hallucinations. |
| Approach: | They propose an object-aware listwise preference optimization for reducing hallucinations in VLMs . they mask a critical object in an image and interpolate the masked region to form more complete images . |
| Outcome: | The proposed method outperforms existing methods in reducing hallucinations and enhancing performance on MMHalBench, AMBER, and Object HalBench. |
Predicting Through Generation: Why Generation Is Better for Prediction (2025.acl-long)
Copied to clipboard
Md Kowsher, Nusrat Jahan Prottasha, Prakash Bhat, Chun-Nam Yu, Mojtaba Soltanalian, Ivan Garibay, Ozlem Garibay, Chen Chen, Niloofar Yousefi
| Challenge: | Large Language Models (LLMs) are increasingly used for predictive tasks such as classification and regression. |
| Approach: | They propose a framework that generates output tokens from mas-sive text corpora and a task adapter to ensure consistency between token generation and final prediction. |
| Outcome: | The proposed framework outperforms baseline models on classification and regression benchmarks and the proposed framework consistently outperformed standard baseline models. |
“Give Me BF16 or Give Me Death”? Accuracy-Performance Trade-Offs in LLM Quantization (2025.acl-long)
Copied to clipboard
| Challenge: | despite popularity of large language model quantization, there are significant accuracy-performance trade-offs associated with quantization formats. |
| Approach: | They evaluate popular quantization formats across academic benchmarks and real-world tasks . they also examine the difference in text generated by quantized models versus their uncompressed counterparts . |
| Outcome: | The proposed format is lossless across all model scales and incurs low accuracy degradation when properly tuned. |
StitchLLM: Serving LLMs, One Block at a Time (2025.acl-long)
Copied to clipboard
Bodun Hu, Shuozhe Li, Saurabh Agarwal, Myungjin Lee, Akshay Jajoo, Jiamin Li, Le Xu, Geon-Woo Kim, Donghyun Kim, Hong Xu, Amy Zhang, Aditya Akella
| Challenge: | Existing techniques like distillation and pruning are not efficient for large language models. |
| Approach: | They propose a dynamic model routing framework that uses a powerful bottom model to process all queries and a lightweight routing mechanism to allocate computational resources appropriately. |
| Outcome: | The proposed framework improves system throughput while minimizing performance degradation. |
Walk in Others’ Shoes with a Single Glance: Human-Centric Visual Grounding with Top-View Perspective Transformation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing VLMs are insensitive to information differences induced by slight perspective changes. |
| Approach: | They propose a visual perspective-taking task that requires robots to interpret human-centric instructions and identify corresponding objects from robot perspectives. |
| Outcome: | The proposed method improves performance by up to 18% and generalizes effectively to robotic and dynamic scenarios. |
Is linguistically-motivated data augmentation worth it? (2025.acl-long)
Copied to clipboard
| Challenge: | Data augmentation is a widely-used technique for addressing data scarcity. |
| Approach: | They compare linguistically-motivated and linguisticly-naive data augmentation strategies for two low-resource languages with different morphological properties. |
| Outcome: | The proposed methods produce synthetic data that follows all linguistic constraints, but they require linguistic expertise and are more difficult to implement. |
From Lists to Emojis: How Format Bias Affects Model Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Format biases in reinforcement learning from human feedback are underexplored . despite its effectiveness, RLHF faces challenges, including policy and regulatory constraints . |
| Approach: | They extend the study of preference biases beyond verbosity bias to a wider range of format biase . they show that with a small amount of biased data, they can inject significant bias into the reward model . |
| Outcome: | The proposed approach can be easily exploited by large language models to achieve higher rankings on popular benchmarks like AlpacaEval and LMSYS Chatbot Arena. |
Colloquial Singaporean English Style Transfer with Fine-Grained Explainable Control (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for style transfer between Singlish and Standard English lack explainability and fine-grained control. |
| Approach: | They propose a multi-agent framework where large language models act as expert agents for each linguistic aspect. |
| Outcome: | The proposed model enables precise, interpretable transformations, advancing explainability in NLP for Singlish. |
From Informal to Formal – Incorporating and Evaluating LLMs on Natural Language Requirements to Verifiable Formal Proofs (2025.acl-long)
Copied to clipboard
Jialun Cao, Yaojie Lu, Meiziniu Li, Haoyang Ma, Haokun Li, Mengda He, Cheng Wen, Le Sun, Hongyu Zhang, Shengchao Qin, Shing-Chi Cheung, Cong Tian
| Challenge: | Recent studies in formal mathematical reasoning have shown an unstoppable growth trend. |
| Approach: | They constructed 18k high-quality instruction-response pairs across five mainstream formal specification languages and evaluated them against ten open-sourced LLMs. |
| Outcome: | The proposed model compared instruction-response pairs across five formal specification languages and found that the LLMs were good at writing proof segments when given either the code, or the detailed description of proof steps. |
CoAM: Corpus of All-Type Multiword Expressions (2025.acl-long)
Copied to clipboard
Yusuke Ide, Joshua Tanner, Adam Nohejl, Jacob Hoffman, Justin Vasselli, Hidetaka Kamigaito, Taro Watanabe
| Challenge: | Existing datasets for multiword expressions are inconsistently annotated, limited to a single type of MWE, or limited in size. |
| Approach: | They propose to use a new interface to generate MWE annotations for the first time in a dataset of MWE identification. |
| Outcome: | The proposed model outperforms existing models on the DiMSUM dataset. |
SeaKR: Self-aware Knowledge Retrieval for Adaptive Retrieval Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Adaptive Retrieval-Augmented Generation (RAG) is an effective strategy to alleviate hallucination of large language models (LLMs). |
| Approach: | They propose a novel adaptive RAG model that extracts self-aware uncertainty of large language models from their internal states and invokes retrieval accordingly. |
| Outcome: | The proposed model outperforms existing adaptive RAG methods on complex and simple Question Answering datasets. |
Exposing the Achilles’ Heel: Evaluating LLMs Ability to Handle Mistakes in Mathematical Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluations focus on final accuracy, neglecting the critical aspect of reasoning capabilities. |
| Approach: | They propose to evaluate LLMs’ abilities to detect and correct reasoning mistakes by using rule-based methods and smaller language models. |
| Outcome: | The proposed model outperforms existing models such as GPT-4o and GPT4 in both accuracy and accuracy, but lacks data contamination and memorization concerns. |
Understanding the Dark Side of LLMs’ Intrinsic Self-Correction (2025.acl-long)
Copied to clipboard
Qingjie Zhang, Di Wang, Haoting Qian, Yiming Li, Tianwei Zhang, Minlie Huang, Ke Xu, Hewu Li, Liu Yan, Han Qiu
| Challenge: | Recent studies show that LLMs’ intrinsic self-correction fails without oracle labels as feedback. |
| Approach: | They propose to use one simple task and three complex tasks with state-of-the-art LLMs like ChatGPT, Llama, and DeepSeek to interpret LLM's intrinsic self-correction. |
| Outcome: | The proposed methods reveal the dark side of LLMs’ intrinsic self-correction for different tasks, especially for those failure cases. |
VideoVista-CulturalLingo: 360° Horizons-Bridging Cultures, Languages, and Domains in Video Comprehension (2025.acl-long)
Copied to clipboard
| Challenge: | Existing video evaluation benchmarks focus on a single language, typically English, and feature videos rooted in Western cultural contexts. |
| Approach: | They propose a video evaluation benchmark designed to bridge cultural, linguistic, and domain divide in video comprehension. |
| Outcome: | The proposed video evaluation benchmark bridges cultural, linguistic, and domain divides . existing benchmarks only feature videos from YouTube, Shutterstock, or established video datasets based on cultural diversity . |
What are the Essential Factors in Crafting Effective Long Context Multi-Hop Instruction Datasets? Insights and Best Practices (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to generate long-context instruction-tuning data are limited by poor quality and fewer than 35% of samples are multi-hop . |
| Approach: | They propose a framework that integrates a quality verification agent, a single-hop question generation agent, and a multi-hop questions merger agent to enhance model performance. |
| Outcome: | The proposed framework significantly improves data quality with high-quality, multi-hop, and diverse data. |
Knowledge Graph Retrieval-Augmented Generation for LLM-based Recommendation (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have produced significant advances in the field of recommender systems. |
| Approach: | They propose to retrieve up-to-date structure information from the knowledge graph to augment recommendations by leveraging external knowledge sources. |
| Outcome: | Experiments on a large dataset show that the proposed method is effective in enhancing LLM-based recommendations. |
SudoLM: Learning Access Control of Parametric Knowledge with Authorization Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing preference alignment is a one-size-fits-all alignment mechanism, where the part of the large language model parametric knowledge with non-preferred features is uniformly blocked to all the users. |
| Approach: | They propose a framework that lets LLMs learn access control over parametric knowledge for users with different credentials via authorization alignment. |
| Outcome: | Experiments on two application scenarios show that the proposed framework effectively controls the user’s access to parametric knowledge and maintains its general utility. |
I0T: Embedding Standardization Method Towards Zero Modality Gap (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies on Contrastive Language-Image Pretraining suffer from a *modality gap* . modality gap occurs when image and text embeddings are projected to disparate manifolds . |
| Approach: | They propose a framework that reduces the modality gap by adding two normalization layers to each encoder. |
| Outcome: | The proposed framework reduces the modality gap while preserving the original embedding representations of trained models with their locked parameters. |
Odysseus Navigates the Sirens’ Song: Dynamic Focus Decoding for Factual and Diverse Open-Ended Text Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing decoding methods struggle to balance factuality and diversity . Deterministic decoding approaches suffer from degeneration and lack of diversity - a problem that is not addressed by the current literature. |
| Approach: | They propose a plug-and-play stochastic approach that adjusts decoding focus based on distributional differences across layers, leveraging the modular nature of factual knowledge within LLMs. |
| Outcome: | Extensive experiments on seven datasets show that DFD significantly improves performance. |
Better Embeddings with Coupled Adam (2025.acl-long)
Copied to clipboard
| Challenge: | Anisotropic Embeddings Large Language Models exhibit undesirable yet poorly understood feature of anisotropy. |
| Approach: | They propose an algorithm that uses the second moment in Adam to mitigate anisotropic embeddings . they propose an embeddable matrix and unembedding matrix to map the input and output tokens based on weight tying . |
| Outcome: | The proposed model improves quality and performance on large datasets. |
Bone Soups: A Seek-and-Soup Model Merging Approach for Controllable Multi-Objective Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches focus on merging language models tuned on single objectives . existing approaches ignore the impacts of competing objectives on model tuning . |
| Approach: | They propose a model merging approach that seeks a series of backbone models and merges them according to user preferences. |
| Outcome: | The proposed approach exhibits strong controllability and Pareto optimality in controllable multi-objective generation. |
Controllable and Reliable Knowledge-Intensive Task-Oriented Conversational Agents with Declarative Genie Worksheets (2025.acl-long)
Copied to clipboard
| Challenge: | Existing LLMs suffer from hallucination, following instructions with conditional logic, and integrating knowledge from different sources. |
| Approach: | They propose a programmable framework for creating knowledge-intensive task-oriented conversational agents that handle involved interactions and answer complex queries. |
| Outcome: | The proposed framework outperforms SOTA methods on complex logic dialogue datasets by up to 20.5%. |
Benchmarking Long-Context Language Models on Long Code Understanding (2025.acl-long)
Copied to clipboard
Jia Li, Xuyuan Guo, Lei Li, Kechi Zhang, Ge Li, Jia Li, Zhengwei Tao, Fang Liu, Chongyang Tao, Yuqi Zhu, Zhi Jin
| Challenge: | Currently, long-context language models are limited by the lack of a rigorous evaluation framework for long code understanding. |
| Approach: | They propose to use a long code understanding benchmark LongCodeU to evaluate LCLMs' long code comprehension ability for practical applications. |
| Outcome: | The proposed benchmarks show that current LCLMs are limited in their long code understanding ability, particularly when the long code length is greater than 32K, falling far short of their claimed 128K to 1M context windows. |
MAGNET: Augmenting Generative Decoders with Representation Learning and Infilling Capabilities (2025.acl-long)
Copied to clipboard
| Challenge: | Decoder-only large language models are increasingly being adapted for bidirectional modeling . however, their reliance on causal attention restricts their effectiveness in tasks that require understanding of bidirectional context. |
| Approach: | They propose a method to adapt decoder-only large language models to generate robust representations and infill missing text spans. |
| Outcome: | The proposed method surpasses strong decoders on token-level and sentence-level representation learning tasks and generates contextually appropriate text infills without excessive repetition of words or phrases. |
Internal Value Alignment in Large Language Models through Controlled Value Vector Activation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing LLMs do not possess consistent values, but many have been developed to align them at the behavioral level, including supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF). |
| Approach: | They propose a Controlled Value Vector Activation method that directly aligns the internal values of Large Language Models by interpreting how a value is encoded in their latent representations. |
| Outcome: | The proposed method achieves highest success rate across 10 basic values without hurting model performance and fluency, and ensures target values even with opposite and potentially malicious input prompts. |
A Dual-Perspective NLG Meta-Evaluation Framework with Automatic Benchmark and Better Interpretability (2025.acl-long)
Copied to clipboard
| Challenge: | Existing evaluation metrics are insufficient to meet requirements for natural language generation. |
| Approach: | They propose a dual-perspective NLG meta-evaluation framework that focuses on different evaluation capabilities and a method of automatically constructing benchmarks without requiring new human annotations. |
| Outcome: | The proposed framework improves interpretability and provides better performance for 16 representative LLMs. |
Recurrent Knowledge Identification and Fusion for Language Model Continual Learning (2025.acl-long)
Copied to clipboard
Yujie Feng, Xujia Wang, Zexin Lu, Shenghong Fu, Guangyuan Shi, Yongxin Xu, Yasha Wang, Philip S. Yu, Xu Chu, Xiao-Ming Wu
| Challenge: | Continual learning (CL) is crucial for large language models without costly retraining. |
| Approach: | They propose a framework for recurrent knowledge identification and fusion that enables dynamic estimation of parameter importance distributions to enhance knowledge transfer. |
| Outcome: | The proposed framework mitigates catastrophic forgetting and enhances knowledge transfer. |
Data-Constrained Synthesis of Training Data for De-Identification (2025.acl-long)
Copied to clipboard
| Challenge: | sensitive domains lack widely available datasets due to privacy risks . recent studies have focused on evaluating the privacy of the synthetic text . |
| Approach: | They domain-adapt LLMs to clinical domain and generate synthetic clinical texts . they then generate NER models that can be annotated with tags for PII . |
| Outcome: | The proposed model performs better than the original model using smaller datasets. |
Just a Scratch: Enhancing LLM Capabilities for Self-harm Detection through Intent Differentiation and Emoji Interpretation (2025.acl-long)
Copied to clipboard
| Challenge: | Self-harm detection on social media is critical for early intervention and mental health support, yet remains challenging due to the subtle, context-dependent nature of such expressions. |
| Approach: | They propose a framework to distinguish intent through nuanced language–emoji interplay. |
| Outcome: | The proposed framework improves self-harm detection and explanation tasks on three state-of-the-art LLMs. |
Contrastive Learning on LLM Back Generation Treebank for Cross-domain Constituency Parsing (2025.acl-long)
Copied to clipboard
| Challenge: | Existing constituency treebanks are limited in out-of-domain settings, therefore constituency parsing is still a challenge. |
| Approach: | They propose a novel method for constituency parsing using large language models . they use a cross-domain constituency treebank to fill missing words with the incomplete one . |
| Outcome: | The proposed method achieves state-of-the-art performance on average compared with baselines on five target domains of MCTB. |
MMDEND: Dendrite-Inspired Multi-Branch Multi-Compartment Parallel Spiking Neuron for Sequence Modeling (2025.acl-long)
Copied to clipboard
Kexin Wang, Yuhong Chou, Di Shang, Shijie Mei, Jiahong Zhang, Yanbin Huang, Man Yao, Bo Xu, Guoqi Li
| Challenge: | Vanilla spiking neurons are simplified from complex biological neurons with dendrites, soma, and synapses into single somatic compartments. |
| Approach: | They propose a multi-branch, multi-compartment parallel spiking dendritic neuron with a proportion-adjustable multi-branched structure that enables long-term temporal dependencies. |
| Outcome: | The proposed model achieves better long-sequence modeling capability with fewer parameters and lower energy consumption. |
Understanding Impact of Human Feedback via Influence Functions (2025.acl-long)
Copied to clipboard
| Challenge: | In reinforcement learning from human feedback, human feedback can be noisy, inconsistent or biased . this variability can lead to misaligned reward signals, potentially causing unintended side effects . |
| Approach: | They propose an approximation method that measures the impact of human feedback on the performance of reward models. |
| Outcome: | The proposed method detects common labeler biases in human feedback datasets and guides labelers in refining their strategies to better align with expert feedback. |
T2I-FactualBench: Benchmarking the Factuality of Text-to-Image Models with Knowledge-Intensive Concepts (2025.acl-long)
Copied to clipboard
Ziwei Huang, Wanggui He, Quanyu Long, Yandi Wang, Haoyuan Li, Zhelun Yu, Fangxun Shu, Weilong Dai, Hao Jiang, Fei Wu, Leilei Gan
| Challenge: | Existing studies on text-to-image (T2I) models focus on text alignment, image quality, and object composition capabilities. |
| Approach: | They propose a T2I-FactualBench benchmark to evaluate the factuality of knowledge-intensive concept generation. |
| Outcome: | The proposed framework evaluates the factuality of knowledge-intensive concept generation tasks. |
InspireDebate: Multi-Dimensional Subjective-Objective Evaluation-Guided Reasoning and Optimization for Debating (2025.acl-long)
Copied to clipboard
| Challenge: | Existing LLMs focus on responding to specific arguments while neglecting objective assessments such as authenticity and logical validity. |
| Approach: | They propose a multi-dimensional evaluation system and an optimized debating framework . they propose to use coT reasoning enhancement, web-based Retrieval Augmented Generation to optimize across various dimensions. |
| Outcome: | The proposed framework outperforms baseline models in argument quality assessment and debate process simulation by 57%. |
OpenWebVoyager: Building Multimodal Web Agents via Iterative Real-World Exploration, Feedback and Optimization (2025.acl-long)
Copied to clipboard
Hongliang He, Wenlin Yao, Kaixin Ma, Wenhao Yu, Hongming Zhang, Tianqing Fang, Zhenzhong Lan, Dong Yu
| Challenge: | Existing studies focus on building text-only agents in synthetic environments where the reward signals are clearly defined. |
| Approach: | They propose a multimodal web agent that can autonomously conduct real-world exploration and improve itself after each iteration. |
| Outcome: | The proposed agent improves itself after each iteration, demonstrating strong performance across multiple test sets. |
FOCUS: Evaluating Pre-trained Vision-Language Models on Underspecification Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | a new dataset evaluates whether vision-language models have underspecification reasoning abilities . underspecifications are often left incomplete or vague, and are often ignored for mutual understanding . |
| Approach: | They propose a probing dataset to evaluate whether VLMs have underspecification reasoning . they find that pre-trained vision-language models lack this ability . |
| Outcome: | The proposed probing dataset shows that pre-trained vision-language models lack underspecification reasoning abilities. |
Sightation Counts: Leveraging Sighted User Feedback in Building a BLV-aligned Dataset of Diagram Descriptions (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that direct generation of diagram descriptions is costly and biased against blind and low-vision (BLV) users. |
| Approach: | They ask sighted individuals to assess diagram descriptions generated by vision-language models . they use latent supervision to guide the models with latent inference . |
| Outcome: | The results show that visual descriptions generated by vision-language models are effective and useful to educators who are themselves BLV and teach visually impaired learners. |
Personal Travel Solver: A Preference-Driven LLM-Solver System for Travel Planning (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) have demonstrated remarkable capabilities in understanding complex instructions and reasoning across diverse domains. |
| Approach: | They propose to integrate user’s implicit preference into the progress of travel planning by integrating real user reviews and point-of-interest metadata from Google Local into RealTravel. |
| Outcome: | The proposed system achieves better performance than baseline methods and improves the level of personalization. |
Counterspeech the ultimate shield! Multi-Conditioned Counterspeech Generation through Attributed Prefix Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to generate counterspeech based on intents are limited to single attributed . however, a holistic approach that considers multiple attributes simultaneously yields more nuanced and effective responses. |
| Approach: | They propose a framework that leverages hierarchical prefix learning with preference optimization to generate more constructive counterspeech. |
| Outcome: | The proposed framework improves intent conformity and emotion labels in 13,973 counterspeech instances. |
LLM×MapReduce: Simplified Long-Sequence Processing using Large Language Models (2025.acl-long)
Copied to clipboard
Zihan Zhou, Chong Li, Xinyi Chen, Shuo Wang, Yu Chao, Zhili Li, Haoyu Wang, Qi Shi, Zhixing Tan, Xu Han, Xiaodong Shi, Zhiyuan Liu, Maosong Sun
| Challenge: | Existing studies have focused on extending the context length of large language models (LLMs) due to their quadratic computational complexity and a lack of high-quality long training examples, most LLMs are trained with a limited window size. |
| Approach: | They propose a training-free framework that enables large language models to effectively process long texts using a divide-and-conquer strategy for comprehensive document understanding. |
| Outcome: | The proposed framework outperforms open-source and commercial long-context LLMs and is compatible with several models. |
CheXalign: Preference fine-tuning in chest X-ray interpretation models without human feedback (2025.acl-long)
Copied to clipboard
Dennis Hein, Zhihong Chen, Sophie Ostmeier, Justin Xu, Maya Varma, Eduardo Pontes Reis, Arne Edward Michalson Md, Christian Bluethgen, Hyun Joo Shin, Curtis Langlotz, Akshay S Chaudhari
| Challenge: | Radiologists are a crucial role in translating medical images into actionable reports . however, the field faces staffing shortages and increasing workloads . |
| Approach: | They propose an automated pipeline for preference feedback focusing on chest X-ray radiology report generation (RRG) method leverages publicly available datasets containing pairs of images and radiologist-written reference reports with reference-based metrics, or Judges. |
| Outcome: | The proposed pipeline achieves state-of-the-art CheXbert scores on the MIMIC-CXR dataset while on average maintaining robust performance across six additional image perception and reasoning tasks. |
Knowledge Tracing in Programming Education Integrating Students’ Questions (2025.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge tracing models that ignore student questions are suboptimal for programming education because of the complexity of coding tasks and the diverse methods students use to solve problems. |
| Approach: | They propose a model that leverages students' questions and automatically extracted skill information to enhance the accuracy of predicting students' performance on subsequent problems. |
| Outcome: | The proposed model improves student performance across Python programming courses of differing difficulty levels. |
PRISM: A Framework for Producing Interpretable Political Bias Embeddings with Political-Aware Cross-Encoder (2025.acl-long)
Copied to clipboard
| Challenge: | Existing embedding models excel at capturing general meaning, but overlook ideological nuances, limiting their effectiveness in political bias tasks. |
| Approach: | They propose a framework to Produce inteRpretable polItical biaS eMbeddings. |
| Outcome: | The proposed framework outperforms state-of-the-art embedding models in political bias classification . the proposed framework offers highly interpretable representations for political analysis . |
Representations of Fact, Fiction and Forecast in Large Language Models: Epistemics and Attitudes (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models estimate and calibrate confidence of large language models with verbalized uncertainty, but they lack a careful examination of the linguistic knowledge of uncertainty encoded in the latent space of LLMs. |
| Approach: | They draw on typological frameworks of epistemic expressions to evaluate LLMs’ knowledge of epistenetic modality, using controlled stories. |
| Outcome: | The proposed models generate expressions matching the strength of evidence and are not robust in generating epistemic expressions. |
Lexical Diversity-aware Relevance Assessment for Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Extensive experiments on widely used benchmarks demonstrate the efficacy of our approach, yielding a 10.6% accuracy improvement on HotpotQA. |
| Approach: | They propose a Lexical Diversity-aware RAG method to address the biases in relevant information retrieval and utilization induced by lexical diversity. |
| Outcome: | Extensive experiments on widely used benchmarks show the proposed method yields a 10.6% accuracy improvement on HotpotQA. |
Weaving Context Across Images: Improving Vision-Language Models through Focus-Centric Visual Chains (2025.acl-long)
Copied to clipboard
| Challenge: | Existing vision-language models struggle to disentangle information scattered across complex visual inputs, leading to performance degradation. |
| Approach: | They propose a focus-centric visual chain paradigm that enhances VLMs’ perception, comprehension, and reasoning abilities in multi-image scenarios. |
| Outcome: | The proposed approach achieves average performance gains of 3.16% and 2.24% across two distinct model architectures, without compromising the general vision-language capabilities. |
Online Iterative Self-Alignment for Radiology Report Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for RRG rely on supervised fine-tuning based on data pairs of radiological images and corresponding radiologist-annotated reports. |
| Approach: | They propose a method that performs supervised fine-tuning on data pairs of radiological images and corresponding radiologist-annotated reports. |
| Outcome: | The proposed method surpasses existing methods and achieves state-of-the-art performance across multiple evaluation metrics. |
Chinese Inertial GAN for Handwriting Signal Generation and Recognition (2025.acl-long)
Copied to clipboard
| Challenge: | Inertial sensors can measure the acceleration and angular velocity of moving objects and are widely used in electronic devices such as smartphones, smartwatches, and fitness bands. |
| Approach: | They propose to use Chinese glyph encoding, forced optimal transport, and semantic relevance alignment to acquire unlimited training samples for Chinese inertial writing recognition. |
| Outcome: | The proposed system improves the performance of six widely used classifiers from 6.7% to 98.4%. |
LLMs Caught in the Crossfire: Malware Requests and Jailbreak Challenges (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have a high vulnerability to jailbreak attacks that leverage crafted prompts to generate malicious outputs. |
| Approach: | They propose to use large language models to test their security against jailbreak attacks that leverage crafted prompts to generate malicious outputs. |
| Outcome: | The proposed model is based on 320 manually crafted malicious code generation requirements, covering 11 jailbreak methods and 29 code functionality categories. |
Evaluating Sequence Labeling on the basis of Information Theory (2025.acl-long)
Copied to clipboard
| Challenge: | Existing metric families focus on certain aspects of sequence labeling tasks. |
| Approach: | They propose a metric that measures how much information each token contributes depending on different aspects of the sequence. |
| Outcome: | The proposed metric can satisfy all properties simultaneously. |
GRAT: Guiding Retrieval-Augmented Reasoning through Process Rewards Tree Search (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to enhance large models for multi-hop question-answering lack the ability for multipath exploration, strategic look-ahead, stepwise evaluation, and global selection. |
| Approach: | They propose an algorithm guided by Monte Carlo Tree Search and process rewards that assigns fine-grained rewards to each step in the search path. |
| Outcome: | The proposed algorithm outperforms various RAG-based methods on four multihop QA datasets and shows that it can self-train and self-update. |
T-REG: Preference Optimization with Token-Level Reward Regularization (2025.acl-long)
Copied to clipboard
| Challenge: | Reinforcement learning from human feedback (RLHF) is a dominant approach for large language models to follow instructions and produce meaningful alignment. |
| Approach: | They propose a method that leverages human feedback to optimize large language models . they propose to use sequence-level and token-level rewards to optimize preference . |
| Outcome: | The proposed method outperforms baseline methods on Alpaca Eval 2 and Arena-Hard benchmarks. |
Gödel Agent: A Self-Referential Agent Framework for Recursively Self-Improvement (2025.acl-long)
Copied to clipboard
| Challenge: | Existing agentic systems cannot search the whole design space due to the restriction of human-designed components. |
| Approach: | They propose a Gödel Agent framework that allows agents to recursively improve themselves without relying on fixed algorithms or fixed algorithms. |
| Outcome: | The proposed framework surpasses manual crafted agents in performance, efficiency, and generalizability. |
AgentGym: Evaluating and Training Large Language Model-based Agents across Diverse Environments (2025.acl-long)
Copied to clipboard
Zhiheng Xi, Yiwen Ding, Wenxiang Chen, Boyang Hong, Honglin Guo, Junzhe Wang, Xin Guo, Dingwen Yang, Chenyang Liao, Wei He, Songyang Gao, Lu Chen, Rui Zheng, Yicheng Zou, Tao Gui, Qi Zhang, Xipeng Qiu, Xuanjing Huang, Zuxuan Wu, Yu-Gang Jiang
| Challenge: | Large language models (LLMs) are promising foundations to build generally-capable agents . however, the community lacks a unified interactive framework that covers diverse environments for comprehensive evaluation of agents. |
| Approach: | They propose a framework that features 7 real-world scenarios, 14 environments, and 89 tasks for unified, real-time, and concurrent agent interaction. |
| Outcome: | The proposed framework features 7 real-world scenarios, 14 environments, and 89 tasks for unified, real-time, and concurrent agent interaction. |
Rethinking the Role of Prompting Strategies in LLM Test-Time Scaling: A Perspective of Probability Theory (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have shown that scaling test-time compute can also effectively improve reasoning. |
| Approach: | They propose a probabilistic method to efficiently predict scaling performance and identify the best prompting strategy under large sampling times. |
| Outcome: | The proposed method significantly improves the scaling performance of majority voting on large language models. |
Information Locality as an Inductive Bias for Neural Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Inductive biases are inherent in every machine learning system, argues a new study . m-local entropy measures how well symbols disambiguate the next symbol . |
| Approach: | They propose a framework that captures local uncertainty of a language by quantifying how effectively preceding symbols disambiguate the next symbol. |
| Outcome: | The proposed framework captures the local uncertainty of a language by quantifying how effectively symbols disambiguate the next symbol. |
Learning to Reason Over Time: Timeline Self-Reflection for Improved Temporal Reasoning in Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models struggle with temporal reasoning, which requires processing time-related information such as event sequencing, durations, and inter-temporal relationships. |
| Approach: | They propose a framework that enhances the temporal reasoning abilities of Large Language Models (LLMs) by combining timeline construction with iterative self-reflection. |
| Outcome: | The proposed framework improves the temporal reasoning abilities of large language models and improves traceability of the inference process. |
Query-driven Document-level Scientific Evidence Extraction from Biomedical Studies (2025.acl-long)
Copied to clipboard
Massimiliano Pronesti, Joao H Bettencourt-Silva, Paul Flanagan, Alessandra Pascale, Oisín Redmond, Anya Belz, Yufang Hou
| Challenge: | Systematic reviews are widely regarded as the gold standard in evidence-based medicine, heavily influencing medical decisions made by doctors, health authorities, and patients. |
| Approach: | They propose a retrieval-augmented generation framework to tackle the unique challenges of evidence extraction by leveraging forest plots from Cochrane systematic reviews. |
| Outcome: | The proposed framework outperforms existing methods by up to 10.3% in the F1 score on this task. |
Towards Robust Universal Information Extraction: Dataset, Evaluation, and Solution (2025.acl-long)
Copied to clipboard
| Challenge: | Existing robust benchmark datasets generate only a limited range of perturbations for a single Information Extraction (UIE) task, which fails to evaluate the robustness of UIE models effectively. |
| Approach: | They propose a new benchmark dataset that utilizes Large Language Models to generate more diverse and realistic perturbations across different IE tasks. |
| Outcome: | The proposed model performs better with only 15% of the data and is more robust with other models. |
Multi-perspective Alignment for Increasing Naturalness in Neural Machine Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Neural machine translation systems amplify lexical biases, rendering outputs artificially impoverished . Attempts to increase naturalness in NMT can fall short in terms of content preservation . |
| Approach: | They propose a method that rewards both naturalness and content preservation . they use multiple perspectives to produce more natural translations . |
| Outcome: | The proposed method produces translations that are lexically richer and exhibit more properties of human-written language without loss in translation accuracy. |
Temporal reasoning for timeline summarisation in social media (2025.acl-long)
Copied to clipboard
| Challenge: | Existing temporal reasoning datasets focus on pair-wise event relationships. |
| Approach: | They propose a temporal reasoning dataset focused on temporal relationships among sequential events within narratives that combines temporal thinking with timeline summarisation through a knowledge distillation framework. |
| Outcome: | The proposed model achieves superior performance on mental health-related timeline summarisation tasks, highlighting the importance and generalisability of leveraging temporal reasoning to improve timeline summaries. |
Beyond Negative Stereotypes – Non-Negative Abusive Utterances about Identity Groups and Their Semantic Variants (2025.acl-long)
Copied to clipboard
| Challenge: | implicitly abusive language is a language that could offend, demean or marginalize another person . a large portion of what is considered abusive language can be classified as implicitly abused . |
| Approach: | They propose to profile implicitly abusive language and use it to analyze a dataset of such utterances. |
| Outcome: | The proposed dataset identifies the type of abusive language that is not conveyed by unambiguously abusive words. |
Persistent Homology of Topic Networks for the Prediction of Reader Curiosity (2025.acl-long)
Copied to clipboard
Manuel D.s. Hopp, Vincent Labatut, Arthur Amalvy, Richard Dufour, Hannah Stone, Hayley K Jach, Kou Murayama
| Challenge: | Existing approaches to model reader engagement rely on surface-level characteristics and often fail to capture the broader semantic structure, narrative flow, and information gaps that stimulate curiosity. |
| Approach: | They propose a framework that quantifies semantic information gaps within a text's semantic structure by using BERTopic-inspired topic modeling and persistent homology to analyze the evolving topology of a dynamic semantic network derived from text segments. |
| Outcome: | The proposed method significantly improves curiosity prediction compared to baseline models (73% vs. 30% explained deviance) |
Tokenisation is NP-Complete (2025.acl-long)
Copied to clipboard
| Challenge: | Tokenisation is the first step required to use a language model. |
| Approach: | They propose to find tokenisers which maximise the compression of a text by either finding a vocabulary directly or selecting a sequence of merge operations. |
| Outcome: | The proposed tokenisation methods are able to find tokenisers that maximise compression of a text. |
Training Dynamics Underlying Language Model Scaling Laws: Loss Deceleration and Zero-Sum Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Increasing language model size improves cross-entropy loss with power-law behaviour, but scaling laws do not explain how scaling improves loss. |
| Approach: | They find that language models undergo loss deceleration early in training . they attribute loss deceleration to a type of degenerate training dynamics we call zero-sum learning . |
| Outcome: | The proposed scaling improves loss on language models, but degrades loss in other subsets, resulting in bottlenecks. |
Parameter-Aware Contrastive Knowledge Editing: Tracing and Rectifying based on Critical Transmission Paths (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) encode vast amounts of knowledge in their parameters, but the acquired knowledge can be incorrect or outdated over time, necessitating rectification after pre-training. |
| Approach: | They propose a method that captures key information flows that influence model predictions . they propose 'critical transmission paths' to improve model editing . |
| Outcome: | The proposed method improves on two prominent datasets and three widely used LLMs. |
Many Heads Are Better Than One: Improved Scientific Idea Generation by A LLM-Based Multi-Agent System (2025.acl-long)
Copied to clipboard
Haoyang Su, Renqi Chen, Shixiang Tang, Zhenfei Yin, Xinzhe Zheng, Jinzhe Li, Biqing Qi, Qi Wu, Hui Li, Wanli Ouyang, Philip Torr, Bowen Zhou, Nanqing Dong
| Challenge: | Recent AI methods have shown promise in tasks such as hypothesis generation and experimental design, but they fail to replicate the collaborative nature of real-world scientific practices. |
| Approach: | They propose a virtual scientific system that mimics the collaborative nature of scientific research by organizing a team of agents to generate, evaluate, and refine research ideas. |
| Outcome: | The proposed system outperforms the state-of-the-art method in producing new scientific ideas and offers valuable insights to guide future research. |
Inner Thinking Transformer: Leveraging Dynamic Depth Scaling to Foster Adaptive Internal Thinking (2025.acl-long)
Copied to clipboard
Yilong Chen, Junyuan Shang, Zhenyu Zhang, Yanxi Xie, Jiawei Sheng, Tingwen Liu, Shuohuan Wang, Yu Sun, Hua Wu, Haifeng Wang
| Challenge: | Large language models face inherent performance bottlenecks under parameter constraints . challenging tokens induce abrupt gradient spikes across layers, exposing stress points . |
| Approach: | They propose an inner thinking transformer that reimagines layer computations as implicit thinking steps. |
| Outcome: | Empirical results show that ITT outperforms Transformer/Loop variants in 11 benchmarks. |
Document-Level Text Generation with Minimum Bayes Risk Decoding using Optimal Transport (2025.acl-long)
Copied to clipboard
| Challenge: | Document-level text generation tasks are more difficult than sentence-level ones as they require an understanding of longer context to generate high-quality texts. |
| Approach: | They propose a variant of minimum bayes risk decoding that uses a sentence-level utility function to compute the utility of a document. |
| Outcome: | The proposed decoding outperforms the standard MBR in document-level tasks. |
Opt-Out: Investigating Entity-Level Unlearning for Large Language Models via Optimal Transport (2025.acl-long)
Copied to clipboard
| Challenge: | Instruction-following large language models (LLMs) inadvertently disclose private, sensitive information to their users, underscoring the need for machine unlearning techniques to remove selective information from the models. |
| Approach: | They propose an optimal transport-based unlearning method that utilizes the Wasserstein distance from the model’s initial parameters to achieve more effective and fine-grained unlearning. |
| Outcome: | The proposed method surpasses existing methods and establishes a new standard for secure and adaptable LLMs that can accommodate user data removal requests without the need for full retraining. |
Mixture of Small and Large Models for Chinese Spelling Check (2025.acl-long)
Copied to clipboard
| Challenge: | Chinese Spelling Check (CSC) tasks have been developed to correct spelling errors in given sentences . fine-tuned BERT-based models show excellent performance but suffer from edit pattern overfitting . a novel mixture approach that effectively combines small models and LLMs during beam search decoding phase improves accuracy and fluency of LLM. |
| Approach: | They propose a dynamic mixture approach that effectively combines small models and LLMs during beam search decoding phase. |
| Outcome: | The proposed method significantly boosts error correction capabilities, achieving state-of-the-art results across multiple datasets. |
DISC: Plug-and-Play Decoding Intervention with Similarity of Characters for Chinese Spelling Check (2025.acl-long)
Copied to clipboard
Ziheng Qiao, Houquan Zhou, Yumeng Liu, Zhenghua Li, Min Zhang, Bo Zhang, Chen Li, Ji Zhang, Fei Huang
| Challenge: | Chinese spelling check (CSC) tasks require that incorrect characters are usually similar to the correct ones in either phonetics or glyph. |
| Approach: | They propose a plug-and-play decoding intervention with similarity of characters module for Chinese spelling check (CSC) they propose to incorporate phonetic and glyph similarities only during the inference phase. |
| Outcome: | The proposed method significantly improves Chinese spelling check models on benchmarks and on benchmark datasets. |
Causal Estimation of Tokenisation Bias (2025.acl-long)
Copied to clipboard
| Challenge: | Modern language models define probabilities over character-strings, but in practice, it does . Ideally, the choice of the tokeniser should not affect the probability assigned to the underlying character- string. |
| Approach: | They quantify a type of tokenisation bias by framing it as a causal effect and estimating it using the regression discontinuity design. |
| Outcome: | The proposed model can estimate tokenisation bias by comparing subwords around arbitrary cutoff points. |
Value Residual Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing decoder-only transformers fail to preserve initial token-level information in deeper layers. |
| Approach: | They propose a new architecture that incorporates value residual connections in addition to hidden state residuals. |
| Outcome: | The proposed architecture reduces KV cache size by nearly half with only a small performance penalty and can be integrated with other KV-efficient methods. |
SGIC: A Self-Guided Iterative Calibration Framework for RAG (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on retrieval-augmented generation (RAG) focus on extracting relevant documents or refinement of specialized instructions. |
| Approach: | They propose a framework that provides LLMs with specific cues to improve their calibration efficacy . they propose an iterative self-calibration training set that harnesses uncertainty scores . |
| Outcome: | The proposed framework significantly improves performance on both closed-source and open-source LLMs. |
NusaAksara: A Multimodal and Multilingual Benchmark for Preserving Indonesian Indigenous Scripts (2025.acl-long)
Copied to clipboard
Muhammad Farid Adilazuarda, Musa Izzanardi Wijanarko, Lucky Susanto, Khumaisa Nur’aini, Derry Tanti Wijaya, Alham Fikri Aji
| Challenge: | NusaAksara covers 8 scripts across 7 languages, including low-resource languages not commonly seen in NLP benchmarks. |
| Approach: | They propose a benchmark for Indonesian scripts that includes their original scripts and a dataset that includes 8 scripts across 7 languages. |
| Outcome: | The proposed benchmark covers 8 scripts across 7 languages, including low-resource languages not commonly seen in NLP benchmarks. |
LLM-based Rumor Detection via Influence Guided Sample Selection and Game-based Perspective Analysis (2025.acl-long)
Copied to clipboard
Zhiliang Tian, Jingyuan Huang, Zejiang He, Zhen Huang, Menglong Lu, Linbo Qiao, Songzhu Mei, Yijie Wang, Dongsheng Li
| Challenge: | Existing methods for rumor detection on social media are limited by limited modeling capacity and insufficient training corpora. |
| Approach: | They propose an SFT-based rumor detection model with Influence guided Sample selection and Game-based multi-perspective analysis to address these issues. |
| Outcome: | The proposed model outperforms existing SOTA on three datasets. |
Hierarchical-Task-Aware Multi-modal Mixture of Incremental LoRA Experts for Embodied Continual Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing continual learning setups for embodied intelligence focus on executing low-level actions, neglecting the ability to learn high-level planning and multi-level knowledge. |
| Approach: | They propose a Hierarchical Embodied Continual Learning Setups (HEC) that divides the agent’s continual learning process into two layers: high-level instructions and low-level actions. |
| Outcome: | The proposed method reduces the forgetting of old tasks compared to other methods, while orthogonally training the remaining parts. |
SpindleKV: A Novel KV Cache Reduction Method Balancing Both Shallow and Deep Layers (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have impressive capabilities across various fields, but their widespread use is facing a severe and realistic challenge, which is their high demand for GPU memory. |
| Approach: | They propose a KV cache reduction method which balances both shallow and deep layers by using an attention weight based eviction method and a codebook based replacement approach. |
| Outcome: | The proposed method reduces the KV cache for shallower layers while preserving similar or even better model performance. |
Medical Graph RAG: Evidence-based Medical Large Language Model via Graph Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
Junde Wu, Jiayuan Zhu, Yunli Qi, Jingkun Chen, Min Xu, Filippo Menolascina, Yueming Jin, Vicente Grau
| Challenge: | GraphRAG framework is designed to enhance LLMs in generating evidence-based medical responses. |
| Approach: | They propose a graph-based Retrieval-augmented generation framework to enhance LLMs in generating evidence-based medical responses. |
| Outcome: | The proposed framework outperforms state-of-the-art models on 9 medical Q&A benchmarks, 2 health fact-checking datasets, and a long-form generation test set. |
Unifying Uniform and Binary-coding Quantization for Accurate Compression of Large Language Models (2025.acl-long)
Copied to clipboard
Seungcheol Park, Jeongin Bae, Beomseok Kwon, Minjun Kim, Byeongwook Kim, Se Jung Kwon, U Kang, Dongsoo Lee
| Challenge: | Quantization is essential for deploying large language models (LLMs) efficiently since they require expensive computational and memory costs. |
| Approach: | They propose a quantization method that unifies flexible mapping techniques to optimize parameters precisely. |
| Outcome: | The proposed method outperforms existing methods and achieves higher accuracy on GSM8K benchmark. |
Agentic Reasoning: A Streamlined Framework for Enhancing LLM Reasoning with Agentic Tools (2025.acl-long)
Copied to clipboard
| Challenge: | Existing reasoning methods excel in structured domains like math and code, but they are not all effective in knowledge-intensive tasks. |
| Approach: | They introduce a framework that enhances large language model reasoning by integrating external tool-using agents. |
| Outcome: | The proposed framework achieves state-of-the-art among public models and delivers comparable performance to OpenAI Deep Research. |
Probing Relative Interaction and Dynamic Calibration in Multi-modal Entity Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Current methods for multi-modal entity alignment ignore relative interactions between modalities and the accuracy of weights. |
| Approach: | They propose a relative interaction and calibration framework for multi-modal entity alignment that uses attention mechanisms to perceive the uncertainty of the weight for each modality. |
| Outcome: | The proposed framework outperforms baselines across 5 datasets and 23 settings. |
Learn to Memorize: Scalable Continual Learning in Semiparametric Models with Mixture-of-Neighbors Induction Memory (2025.acl-long)
Copied to clipboard
| Challenge: | Semiparametric language models (LMs) use static storage, which lacks learning capability and is disconnected from the internal information flow of the parametric models. |
| Approach: | They reconceptualize the non-parametric memory represented by kNN-LM as a learnable Mixture-of-Neighbors Induction Memory (MoNIM) this synergizes the induction capabilities of attention heads with the memorization strength of feed-forward networks . |
| Outcome: | The proposed model is a learnable Mixture-of-neighbors induction memory (MoNIM) it synergizes the induction capabilities of attention heads with the memorization strength of feed-forward networks (FFNs). |
Adverse Event Extraction from Discharge Summaries: A New Dataset, Annotation Scheme, and Initial Findings (2025.acl-long)
Copied to clipboard
Imane Guellil, Salomé Andres, Atul Anand, Bruce Guthrie, Huayu Zhang, Abul Hasan, Honghan Wu, Beatrice Alex
| Challenge: | Existing resources for AE extraction are limited due to complexity, variability, and ambiguity of clinical narratives. |
| Approach: | They present a manually annotated corpus for Adverse Event (AE) extraction from discharge summaries of elderly patients. |
| Outcome: | The proposed model performs well on coarse-grained extraction, but drops notably for rare events and complex attributes. |
Speed Up Your Code: Progressive Code Acceleration Through Bidirectional Tree Editing (2025.acl-long)
Copied to clipboard
Longhui Zhang, Jiahao Wang, Meishan Zhang, GaoXiong Cao, Ensheng Shi, Mayuchi Mayuchi, Jun Yu, Honghai Liu, Jing Li, Min Zhang
| Challenge: | Existing training methods, such as direct instruction fine-tuning, overlook hierarchical relationships among acceleration patterns. |
| Approach: | They propose a new training paradigm that uses bidirectional tree editing and progressive code acceleration learning to improve LLMs’ CA capabilities. |
| Outcome: | The proposed training paradigm outperforms prompt-enhanced GPT-4 and current training-based methods on average across five programming languages. |
Multi-Facet Blending for Faceted Query-by-Example Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to faceted query-by-example (QBE) rely on document-level comparisons using basic indicators like citations . however, this limited their use to citation-based domains and fails to capture the intricacies of facet constraints. |
| Approach: | They propose a multi-facet blending augmentation method that exploits modularity by decomposing and recomposing to synthesize facet-specific training sets. |
| Outcome: | The proposed method decomposes documents into facet units and generates (ir)relevant pairs, thereby synthesizing facet-specific training sets. |
PIPER: Benchmarking and Prompting Event Reasoning Boundary of LLMs via Debiasing-Distillation Enhanced Tuning (2025.acl-long)
Copied to clipboard
Zhicong Lu, Changyuan Tian, PeiguangLi PeiguangLi, Li Jin, Sirui Wang, Wei Jia, Ying Shen, Guangluan Xu
| Challenge: | Existing studies on Large Language Models (LLMs) have failed to evaluate their performance in event reasoning with a single event relational type or reasoning format. |
| Approach: | They propose a benchmark to evaluate LLMs' event reasoning capability using a single event relational type or reasoning format. |
| Outcome: | The proposed model improves on 10K diverse instruction-tuning demonstrations to alleviate event reasoning-oriented data scarcity. |
MIR: Methodology Inspiration Retrieval for Scientific Research Problems (2025.acl-long)
Copied to clipboard
Aniketh Garikaparthi, Manasi Patwardhan, Aditya Sanjiv Kanade, Aman Hassan, Lovekesh Vig, Arman Cohan
| Challenge: | Existing methods for generating ideas rely on grounding the discovery process within the literature, but their effectiveness varies significantly with the quality and nature of the retrieved literature. |
| Approach: | They construct a methodological inspiration retrieval task using a citation-based methodology adjacency graph and embed an "intuitive prior'' into dense retrievers. |
| Outcome: | The proposed method achieves significant gains in Recall@3 and mAP over strong baselines. |
Sticking to the Mean: Detecting Sticky Tokens in Text Embedding Models (2025.acl-long)
Copied to clipboard
| Challenge: | Sticky tokens, when repeatedly inserted into sentences, pull sentence similarity toward a certain value, disrupting the normal distribution of embedding distances and degrading downstream performance. |
| Approach: | They propose a method to detect “sticky tokens” by sentence and token filtering and apply it to 40 checkpoints across 14 model families. |
| Outcome: | The proposed method detects 868 sticky tokens across 14 models and shows that their presence does not correlate with model size or vocabulary size. |
Memorizing is Not Enough: Deep Knowledge Injection Through Reasoning (2025.acl-long)
Copied to clipboard
Ruoxi Xu, Yunjie Ji, Boxi Cao, Yaojie Lu, Hongyu Lin, Xianpei Han, Ben He, Yingfei Sun, Xiangang Li, Le Sun
| Challenge: | Existing knowledge injection frameworks focus on knowledge memorization and retrieval, but static nature of large language models leads to outdated information as the real world evolves or when adapting to domain-specific knowledge. |
| Approach: | They propose a four-tier knowledge injection framework that defines the levels of knowledge injection: memorization, retrieval, reasoning, and association. |
| Outcome: | The proposed framework defines the levels of knowledge injection: memorization, retrieval, reasoning, and association. |
Improving Dialogue State Tracking through Combinatorial Search for In-Context Examples (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for training dialogue state tracking data are suboptimal . existing methods rely on suboptimized data, resulting in poor performance . |
| Approach: | They propose a method that scores effective in-context examples based on their combinatorial impact on DST performance. |
| Outcome: | The proposed method achieves a 20% gain in data efficiency and generalizing well to the SGD dataset. |
Pretraining Context Compressor for Large Language Models with Embedding-Based Memory (2025.acl-long)
Copied to clipboard
| Challenge: | Efficient processing of long contexts in large language models is essential for real-world applications such as retrieval-augmented generation and in-context learning. |
| Approach: | They propose a decoupled compressor-LLM framework that preserves contextual information within condensed embedding representations. |
| Outcome: | The proposed framework outperforms baseline models in three domains and across eight datasets while adapting to different downstream LLMs. |
Dialogue Systems for Emotional Support via Value Reinforcement (2025.acl-long)
Copied to clipboard
| Challenge: | Emotional support dialogue systems aim to reduce help-seekers’ distress and help them overcome challenges. |
| Approach: | They propose a value-driven method for training emotional support dialogue systems designed to reinforce positive values in seekers by leveraging online support conversations from Reddit. |
| Outcome: | The proposed model outperforms baseline models across support skills, seekers’ emotional intensity, and value reinforcement. |
Length-Induced Embedding Collapse in PLM-based Models (2025.acl-long)
Copied to clipboard
| Challenge: | In text embeddings from PLMs are essential for many NLP applications, but performance degrades on longer texts. |
| Approach: | They propose a method which mitigates the phenomenon of Length Collapse . they propose TempScale to ensure more consistent embeddings across different text lengths . |
| Outcome: | The proposed method improves performance on MTEB and LongEmbed by 0.94% on short and 1.10% on long texts. |
SHuBERT: Self-Supervised Sign Language Representation Learning via Multi-Stream Cluster Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for sign language processing have relied on task-specific models, limiting the potential for transfer learning across tasks. |
| Approach: | They propose a self-supervised contextual representation model that adapts masked token prediction objectives to multi-stream visual sign language input. |
| Outcome: | The proposed model adapts masked token prediction objectives to multi-stream visual sign language input, learning to predict multiple targets corresponding to clustered hand, face, and body pose streams. |
ERU-KG: Efficient Reference-aligned Unsupervised Keyphrase Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for keyphrase prediction rely on heuristicically defined importance scores . existing methods lack consideration for time efficiency . |
| Approach: | They propose an unsupervised keyphrase generation model that combines informativeness and phraseness modules. |
| Outcome: | The proposed model outperforms baseline models and achieves 89% of the performance of a supervised model for top 10 predictions. |
Know Your Mistakes: Towards Preventing Overreliance on Task-Oriented Conversational AI Through Accountability Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Recent LLMs are known to hallucinate, producing responses that seem plausible but are factually incorrect. |
| Approach: | They propose an accountability model for LLM-based task-oriented dialogue agents to address user overreliance via friction turns in cases of model uncertainty and errors associated with dialogue state tracking (DST). |
| Outcome: | The proposed model improves joint goal accuracy (JGA) of DST output by 3% on two established benchmarks. |
LLMs Trust Humans More, That’s a Problem! Unveiling and Mitigating the Authority Bias in Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
Yuxuan Li, Xinwei Guo, Jiashi Gao, Guanhua Chen, Xiangyu Zhao, Jiaxin Zhang, Quanying Liu, Haiyan Wu, Xin Yao, Xuetao Wei
| Challenge: | Large language models (LLMs) generate outputs that stray from user input or contravene established knowledge. |
| Approach: | They propose a new phenomenon, Authority Bias, where LLMs favor one knowledge source over the other . they propose atomic information that generates conflicts and a Conflict Detection Enhanced Query framework . |
| Outcome: | The proposed framework reduces Authority bias in large language models . it detects conflicts, performs credibility assessment on conflicting paragraphs, and detects perturbed text . |
Divide-Then-Aggregate: An Efficient Tool Learning Method via Parallel Tool Invocation (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) demonstrate remarkable capabilities but their ability to autonomously execute complex real-world tasks remains limited. |
| Approach: | They propose a parallel tool invocation framework that decomposes tasks into parallel tool-using subtasks while aggregating results for subsequent decisions. |
| Outcome: | The proposed method significantly improves task performance while reducing token consumption and inference time. |
Reviving Cultural Heritage: A Novel Approach for Comprehensive Historical Document Restoration (2025.acl-long)
Copied to clipboard
Yuyi Zhang, Peirong Zhang, Zhenhua Yang, Pengyu Yan, Yongxin Shi, Pengwei Liu, Fengjun Guo, Lianwen Jin
| Challenge: | Existing methods for historical document restoration focus on single modality or limited-size restoration, failing to meet practical needs. |
| Approach: | They propose a full-page HDR dataset and an automated HDR solution to replace manual restoration methods. |
| Outcome: | The proposed solution improves OCR accuracy from 46.83% to 84.05% when processing severely damaged documents, with enhancement to 94.25% through human-machine collaboration. |
PopAlign: Diversifying Contrasting Patterns for a More Comprehensive Alignment (2025.acl-long)
Copied to clipboard
Zekun Moore Wang, Shenzhi Wang, King Zhu, Jiaheng Liu, Ke Xu, Jie Fu, Wangchunshu Zhou, Wenhao Huang
| Challenge: | Typical approaches to training large language models rely on limited contrasting patterns . contrasting data is limited and models are susceptible to harmful response tendencies . |
| Approach: | They propose a framework that integrates contrasting patterns across the prompt, model, and pipeline levels. |
| Outcome: | The proposed framework outperforms existing methods in the comparison of RQ1 and RQ2 . the proposed framework significantly outperformed existing methods, leading to more comprehensive alignment. |
Robust Utility-Preserving Text Anonymization Based on Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing techniques face challenges of re-identification ability of large language models . anonymizing text that contains sensitive information is crucial for a wide range of applications . |
| Approach: | They propose a framework that integrates three key LLM components to perform anonymization. |
| Outcome: | The proposed model outperforms baselines while maintaining greater data utility in downstream tasks. |
SEAL: Scaling to Emphasize Attention for Long-Context Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies show that advanced LLMs suffer from degradation when processing longer context data. |
| Approach: | They propose a learning-based mechanism that leverages generated data to emphasize attention heads for long-context retrieval. |
| Outcome: | The proposed approach improves retrieval performance over long contexts while maintaining high reliability. |
From Neurons to Semantics: Evaluating Cross-Linguistic Alignment Capabilities of Large Language Models via Neurons Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Existing alignment benchmarks focus on sentence embeddings, but prior research has shown that neural models tend to induce a non-smooth representation space, which impact of semantic alignment evaluation on low-resource languages. |
| Approach: | They propose a novel cross-lingual alignment evaluation method based on the consistency of parallel sentences to assess model alignment. |
| Outcome: | The proposed method achieves a correlation of 0.9556 with downstream tasks performance and 0.8524 with transferability even with a small dataset. |
𝒜3: Automatic Alignment Framework for Attributed Text Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches of aligning large language models to follow user instructions can lead to undue emphasis on irrelevant documents, which in turn reduces the quality of responses. |
| Approach: | They propose to use a framework to automatically generate high-quality attributed query-response pairs for both supervised fine-tuning and preference optimization stages without human annotation. |
| Outcome: | The proposed framework can generate high-quality attributed query-response pairs without human annotation without human intervention. |
Towards Better Value Principles for Large Language Model Alignment: A Systematic Evaluation and Enhancement (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) show remarkable performance across tasks . alignment with human values is critical for their responsible development. |
| Approach: | They propose a framework that evaluates value principles along three desirable properties . they propose supervised fine-tuning, reinforcement learning-based approaches . |
| Outcome: | The proposed framework improves value principles along the three desirable properties of LLMs. |
Language Models, Graph Searching, and Supervision Adulteration: When More Supervision is Less and How to Make More More (2025.acl-long)
Copied to clipboard
| Challenge: | Decoder-only LMs fail to solve the path-star task above 1/D chance due to a learned shortcut that absorbs training supervision. |
| Approach: | They propose a path-star task which is a minimal example of searching over a graph with D arms rooted at a single start node and a query to generate the arm with t from s to t. |
| Outcome: | The proposed task is solvable via decoder-only LMs and its minimal nature prevents its decomposition. |
Diversity Explains Inference Scaling Laws: Through a Case Study of Minimum Bayes Risk Decoding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that MBR decoding improves model generation performance . however, the theoretical underpinnings of these results remain uncertain . |
| Approach: | They propose a theoretical interpretation of MBR decoding from the perspective of bias–diversity decomposition. |
| Outcome: | The proposed method improves the quality estimation of hypotheses by decomposing bias and diversity into two main factors. |
Performance Gap in Entity Knowledge Extraction Across Modalities in Vision Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Vision-language models excel at extracting and reasoning about information from images, yet their capacity to leverage internal knowledge about specific entities remains underexplored. |
| Approach: | They propose a dataset which allows separating entity recognition and question answering . they hypothesize that this decline arises from limitations in how information flows from image tokens to query tokens. |
| Outcome: | The proposed model performance drops when the entity is presented visually rather than textually. |
SDD: Self-Degraded Defense against Malicious Fine-tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Open-source Large Language Models (LLMs) employ safety alignment methods to resist harmful instructions, but malicious fine-tuning can easily bypass these safeguards. |
| Approach: | They propose a framework to prevent malicious fine-tuning of large language models on harmful data by using alignment methods that encourage them to produce irrelevant responses to harmful prompts. |
| Outcome: | The proposed framework reduces the general capability of the LLM when malicious fine-tuning fails, rendering it incapable of following harmful instructions. |
CoachMe: Decoding Sport Elements with a Reference-Based Coaching Instruction Generation Model (2025.acl-long)
Copied to clipboard
Wei-Hsin Yeh, Yu-An Su, Chih-Ning Chen, Yi-Hsueh Lin, Calvin Ku, Wenhsin Chiu, Min-Chun Hu, Lun-Wei Ku
| Challenge: | Existing multimodal models for motion related tasks have shown significant progress. |
| Approach: | They propose a reference-based model that analyzes the differences between a learner’s motion and a physical reference under temporal and physical aspects. |
| Outcome: | The proposed model outperforms GPT-4o on figure skating and boxing by 31.6% and 58.3% respectively. |
DRPruning: Efficient Large Language Model Pruning through Distributionally Robust Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Structured pruning reduces model size but often causes uneven degradation across domains, leading to biased performance. |
| Approach: | They propose a method that dynamically adjusts the data distribution during training to restore balanced performance across heterogeneous and multi-tasking data. |
| Outcome: | Experiments in monolingual and multilingual settings show that the proposed method surpasses similarly sized models in pruning and continued pretraining over perplexity, downstream tasks, and instruction tuning. |
How LLMs Comprehend Temporal Meaning in Narratives: A Case Study in Cognitive Evaluation of LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models exhibit increasingly sophisticated linguistic capabilities, yet the extent to which these models reflect human-like cognition versus advanced pattern recognition remains an open question. |
| Approach: | They conduct a series of targeted experiments to assess whether LLMs construct semantic representations and pragmatic inferences in a human-like manner. |
| Outcome: | The proposed framework can be used to assess the cognitive and linguistic capabilities of large language models (LLMs). |
Data Caricatures: On the Representation of African American Language in Pretraining Corpora (2025.acl-long)
Copied to clipboard
Nicholas Deas, Blake Vente, Amith Ananthram, Jessica A Grieser, Desmond U. Patton, Shana Kleiner, James R. Shepard Iii, Kathleen McKeown
| Challenge: | Recent work in linguistics and NLP has investigated the quantity and quality of AAL representation in pretraining corpora. |
| Approach: | They examine the quantity and quality of African American Language (AAL) representation in pretraining corpora. |
| Outcome: | The results show that AAL is underrepresented in all evaluated corpora compared to US demographics . they also show that most automated filters are more likely to conserve white Mainstream English (WME) texts over AAL . |
Language Model Probabilities are Not Calibrated in Numeric Contexts (2025.acl-long)
Copied to clipboard
Charles Lovering, Michael Krumdick, Viet Dac Lai, Varshini Reddy, Seth Ebner, Nilesh Kumar, Rik Koncel-Kedziorski, Chris Tanner
| Challenge: | Using language model outputs, we find that even in simple settings, the best LMs (1) are poorly calibrated and (2) have systematic biases. |
| Approach: | They argue that language model outputs should capture natural distributions over multiple options within their textual contexts. |
| Outcome: | The proposed model outputs are calibrated to the numeric content of their contexts. |
MDCure: A Scalable Pipeline for Multi-Document Instruction-Following (2025.acl-long)
Copied to clipboard
| Challenge: | Multi-document (MD) processing is crucial for LLMs to handle real-world tasks such as summarization and question-answering across large sets of documents. |
| Approach: | They propose a framework that generates high-quality synthetic MD instruction data over sets of articles via targeted prompts. |
| Outcome: | MDCure generates high-quality synthetic MD instruction data over sets of articles . evaluations show it improves over pre-trained models by up to 75.1% . |
Cross-Lingual Auto Evaluation for Assessing Multilingual LLMs (2025.acl-long)
Copied to clipboard
Sumanth Doddapaneni, Mohammed Safi Ur Rahman Khan, Dilip Venkatesh, Raj Dabre, Anoop Kunchukuttan, Mitesh M Khapra
| Challenge: | Evaluating machine-generated text remains a challenge in NLP for non-English languages . current evaluation frameworks focus on English, revealing a gap in multilingual evaluations . |
| Approach: | They propose a cross-lingual auto evaluation framework that includes evaluator LLMs and a test set specifically designed for multilingual evaluation. |
| Outcome: | The proposed model aligns more closely with human judgments than proprietary models on non-English language evaluations. |
DeepReview: Improving LLM-based Paper Review with Human-like Deep Thinking Process (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Large Language Models (LLMs) face limited domain expertise, hallucinated reasoning, and a lack of structured evaluation. |
| Approach: | They propose a multi-stage framework to emulate expert reviewers by incorporating structured analysis, literature retrieval, and evidence-based argumentation. |
| Outcome: | The proposed model outperforms CycleReviewer-70B with fewer tokens and achieves 88.21% and 80.20% win rates. |
Bypass Back-propagation: Optimization-based Structural Pruning for Large Language Models via Policy Gradient (2025.acl-long)
Copied to clipboard
| Challenge: | Recent pruning methods rely on heuristically hand-crafted metrics, leading to suboptimal performance. |
| Approach: | They propose a method that optimizes pruning masks by minimizing back-propagation . they learn an underlying Bernoulli distribution to sample binary pruning mask samples . |
| Outcome: | The proposed method is able to support global and heterogeneous pruning without back-propagation. |
Tree-of-Debate: Multi-Persona Debate Trees Elicit Critical Thinking for Scientific Comparative Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Existing comparative summarization methods focus on surface-level semantic differences, which may not capture the most relevant distinctions. |
| Approach: | They propose a framework which transforms scientific papers into LLM personas that debate their respective novelties. |
| Outcome: | The proposed framework generates informative arguments and effectively contrasts papers, and supports researchers in their literature review. |
Hierarchical Memory Organization for Wikipedia Generation (2025.acl-long)
Copied to clipboard
Eugene J. Yu, Dawei Zhu, Yifan Song, Xiangyu Wong, Jiebin Zhang, Wenxuan Shi, Xiaoguang Li, Qun Liu, Sujian Li
| Challenge: | Existing methods for generating Wikipedia articles do not utilize memory directly for outline generation. |
| Approach: | They propose a method to generate Wikipedia articles autonomously by leveraging a hierarchical memory architecture. |
| Outcome: | The proposed framework outperforms baseline methods in producing informative and reliable articles. |
Class Distillation with Mahalanobis Contrast: An Efficient Training Paradigm for Pragmatic Language Understanding Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing classifiers for detecting deviant language often come with significant computational cost and high data demands. |
| Approach: | They propose a class-disstillation paradigm that targets the core challenge: distilling a small, well-defined target class from a heterogeneous background. |
| Outcome: | The proposed training paradigm outperforms baselines and large language models on three benchmarks. |
Structure-aware Domain Knowledge Injection for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Structure-aware Continual Pre-Training (SCPT) and Structure-Aware Supervised Fine-Tuning (SSFT) are two-stage strategies for knowledge injection and alignment that reduces the training corpus needs to 5% while achieving 100% of traditional knowledge injection performance. |
| Approach: | They propose a method to efficiently transform foundation Large Language Models into domain specialists by using two-stage strategies: Structure-aware Continual Pre-Training and Structure-Aware Supervised Fine-Tuning. |
| Outcome: | The proposed method significantly reduces the training corpus needs to a mere 5% while achieving 100% of traditional knowledge injection performance. |
FinMME: Benchmark Dataset for Financial Multi-Modal Reasoning Evaluation (2025.acl-long)
Copied to clipboard
Junyu Luo, Zhizhuo Kou, Liming Yang, Xiao Luo, Jinsheng Huang, Zhiping Xiao, Jingshu Peng, Chengzhong Liu, Jiaming Ji, Xuanzhe Liu, Sirui Han, Ming Zhang, Yike Guo
| Challenge: | Multimodal Large Language Models (MLLMs) have experienced rapid development in recent years, but there is a notable lack of effective and specialized multimodal evaluation datasets in the financial domain. |
| Approach: | They introduce FinMME, a multimodal large language model with 11,000 financial research samples and 20 annotators. |
| Outcome: | The proposed model performs better than state-of-the-art models, highlighting its challenging nature. |
Dialectal Coverage And Generalization in Arabic Speech Recognition (2025.acl-long)
Copied to clipboard
| Challenge: | Existing ASR systems cover the modern standard Arabic variety but fail to cover the multitude of spoken variants. |
| Approach: | They propose a suite of automatic speech recognition models optimized to recognize multiple variants of spoken Arabic. |
| Outcome: | The proposed models show coverage and performance gains compared to prior models. |
EditInspector: A Benchmark for Evaluation of Text-Guided Image Edits (2025.acl-long)
Copied to clipboard
| Challenge: | Text-guided image editing is becoming increasingly widespread . current models struggle to evaluate edits comprehensively and often hallucinate when describing changes. |
| Approach: | They propose a novel framework to evaluate edits based on human annotations . they use a template to collect human annotation data and validate the results . |
| Outcome: | The proposed methods outperform current models in artifact detection and difference caption generation. |
Reconsidering LLM Uncertainty Estimation Methods in the Wild (2025.acl-long)
Copied to clipboard
Yavuz Faruk Bakman, Duygu Nur Yaldiz, Sungmin Kang, Tuo Zhang, Baturalp Buyukates, Salman Avestimehr, Sai Praneeth Karimireddy
| Challenge: | Existing studies evaluate UE methods in short-form QA settings, but real-world deployment presents several challenges. |
| Approach: | They examine UE methods' sensitivity to decision threshold selection and their robustness to query transformations such as typos and adversarial prompts. |
| Outcome: | The proposed methods exhibit robustness against typos, adversarial prompts, and prior chat history, and are highly susceptible to adversarials. |
Bregman Conditional Random Fields: Sequence Labeling with Parallelizable Inference Algorithms (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for sequence labeling are hidden Markov models and conditional random fields (CRF). |
| Approach: | They propose a new discriminative model for sequence labeling called Bregman conditional random fields (BCRF) they propose to use Fenchel-Young losses to learn from partial labels. |
| Outcome: | The proposed model performs better in highly constrained settings than the existing model, which is slower and faster. |
SEE: Strategic Exploration and Exploitation for Cohesive In-Context Prompt Optimization (2025.acl-long)
Copied to clipboard
Wendi Cui, Jiaxin Zhang, Zhuohang Li, Hao Sun, Damien Lopez, Kamalika Das, Bradley A. Malin, Sricharan Kumar
| Challenge: | Existing approaches separate the optimization of prompt instructions and in-context learning examples, leading to incohesive, suboptimal results. |
| Approach: | They propose a framework that refines both prompt instructions and in-context learning examples. |
| Outcome: | The proposed framework outperforms state-of-the-art prompt optimization methods on 35 benchmark tasks. |
Programming by Example meets Historical Linguistics: A Large Language Model Based Approach to Sound Law Induction (2025.acl-long)
Copied to clipboard
Atharva Naik, Darsh Agrawal, Hong Sng, Clayton Marr, Kexun Zhang, Nathaniel Romney Robinson, Kalvin Chang, Rebecca Byrnes, Aravind Mysore, Carolyn Rose, David R. Mortensen
| Challenge: | Historical linguists have written programs that convert reconstructed words into their attested descendants via ordered string rewrite functions. |
| Approach: | They propose to use a model to generate a "similar distribution" for sound law induction . they propose four kinds of methods with varying amounts of inductive bias to investigate best performance . |
| Outcome: | The proposed model shows that it can be fine tuned with training data and evaluation data. |
Synergizing Unsupervised Episode Detection with LLMs for Large-Scale News Events (2025.acl-long)
Copied to clipboard
| Challenge: | State-of-the-art automatic event detection struggles with interpretability and adaptability to evolving large-scale key events. |
| Approach: | They propose a task which identifies episodes within a news corpus of key event articles. |
| Outcome: | The proposed framework achieves 59.2% gain across all metrics compared to baselines. |
Beyond True or False: Retrieval-Augmented Hierarchical Analysis of Nuanced Claims (2025.acl-long)
Copied to clipboard
| Challenge: | Claims are often nuanced and cannot be clearly labeled as “true” or “false” . however, a claim can be dissected into integral aspects and sub-aspects that are individually easier to validate . |
| Approach: | They propose a retrieval-augmented generation-based framework for deconstructing nuanced claims . claim can be dissected into integral aspects and sub-aspects, which are easier to validate . |
| Outcome: | The proposed framework can be easily deconstructed into integral aspects and sub-aspects, which are easier to validate. |
The Task Shield: Enforcing Task Alignment to Defend Against Indirect Prompt Injection in LLM Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Model (LLM) agents are becoming conversational assistants . indirect prompt injection attacks pose a critical threat to these systems . |
| Approach: | They propose a novel and orthogonal perspective that reframes agent security . they propose 'task shield' that verifies whether each instruction and tool call contributes to user objectives . |
| Outcome: | The proposed defense reduces attack success rates while maintaining high task utility on the AgentDojo benchmark. |
Sandcastles in the Storm: Revisiting the (Im)possibility of Strong Watermarking (2025.acl-long)
Copied to clipboard
Fabrice Y Harel-Canada, Boran Erol, Connor Choi, Jason Liu, Gary Jiarui Song, Nanyun Peng, Amit Sahai
| Challenge: | Recent theoretical work argues that any watermark can be erased via random walk attacks that perturb text while preserving quality. |
| Approach: | They propose that any watermark can be erased via random walk attacks that perturb text while preserving quality. |
| Outcome: | The proposed method underperforms the theoretical models in large-scale experiments and human-validated assessments. |
Time-MQA: Time Series Multi-Task Question Answering with Context Enhancement (2025.acl-long)
Copied to clipboard
Yaxuan Kong, Yiyuan Yang, Yoontae Hwang, Wenjie Du, Stefan Zohren, Zhangyang Wang, Ming Jin, Qingsong Wen
| Challenge: | Existing time series models focus on a narrow spectrum of tasks, such as forecasting or anomaly detection. |
| Approach: | They propose a framework that enables natural language queries across multiple time series tasks such as numerical analytical tasks and open-ended question answering with reasoning. |
| Outcome: | The proposed framework enables natural language queries across multiple time series tasks and allows for more advanced and intuitive interactions with temporal data. |
From Perceptions to Decisions: Wildfire Evacuation Decision Prediction with Behavioral Theory-informed LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing statistical methods for evacuation decision prediction fail to capture complex and diverse behavioral logic of different individuals. |
| Approach: | They propose a Large Language Model (LLM)-based framework that integrates behavioral theories and models to streamline the Chain-of-Thought reasoning and integrates with memory-based Reinforcement Learning module to provide accurate evacuation decision prediction and understanding. |
| Outcome: | The proposed framework improves on three post-wildfire survey datasets with strong cross-event generalizability over existing models. |
GETReason: Enhancing Image Context Extraction through Hierarchical Multi-Agent Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for analyzing images of events fail to accurately extract contextual meaning from images. |
| Approach: | They propose a framework to infer global event, temporal, and geospatial information from images . they also introduce a new metric GREAT for a reasoning-weighted evaluation . |
| Outcome: | The proposed framework enables better understanding of event significance from images . it also shows that meaningful information can be inferred from images, allowing them to be effectively linked to their corresponding events and contextual background. |
Hanging in the Balance: Pivotal Moments in Crisis Counseling Conversations (2025.acl-long)
Copied to clipboard
| Challenge: | During a conversation, there can come certain moments where its outcome hangs in the balance. |
| Approach: | They propose an unsupervised computational method for detecting pivotal moments as they happen. |
| Outcome: | The proposed method aligns with human perception and the eventual conversational trajectory, which is more likely to change course at these moments. |
Unveiling the Potential of BERT-family: A New Recipe for Building Scalable, General and Competitive Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Generative large language models (LLMs) have significantly influenced various aspects of society, reshaping how we access and interact with information and knowledge. |
| Approach: | They propose a pre-training task that helps BERT-family excel in wider applications . they also explore the integration of cutting-edge technologies into their models to further enhance their capabilities. |
| Outcome: | The proposed model exhibits performance levels comparable to current SOTA LLMs across a spectrum of tasks. |
TaxoAdapt: Aligning LLM-Based Multidimensional Taxonomy Construction to Evolving Research Corpora (2025.acl-long)
Copied to clipboard
| Challenge: | Recent automated taxonomies over-rely on a specific corpus, sacrificing generalizability, or depend heavily on the general knowledge of large language models (LLMs) . |
| Approach: | They propose a framework that dynamically adapts an LLM-generated taxonomy to a given corpus across multiple dimensions. |
| Outcome: | The proposed framework performs iterative hierarchical classification, expanding both the taxonomy width and depth based on corpus’ topical distribution. |
An Empirical Study of Iterative Refinements for Non-autoregressive Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Iterative non-autoregressive (NAR) models have recently demonstrated impressive performance in varied generation tasks, surpassing the autoregressive Transformer. |
| Approach: | They propose a strategy to conduct efficient refinements without performance declines by using two simple metrics to identify potential problems existing in current refinement processes. |
| Outcome: | The proposed model outperforms the autoregressive Transformer by around one BLEU on average. |
Retrofitting Large Language Models with Dynamic Tokenization (2025.acl-long)
Copied to clipboard
| Challenge: | Current language models use a static tokenizer, which results in degraded efficiency and language capabilities. |
| Approach: | They propose a way to dynamically decide on token boundaries based on input text . they merge frequent subword sequences in a batch and apply a hypernetwork to compute token embeddings . |
| Outcome: | The proposed method reduces token sequence lengths by >20% across 14 languages while promoting fairness across languages. |
Principled Content Selection to Generate Diverse and Personalized Multi-Document Summaries (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models exhibit the _”lost in the middle” phenomenon when they are unevenly attending to different parts of the provided context. |
| Approach: | They propose principled content selection as a way to increase source coverage . they use determinantal point processes to prioritize diverse content . |
| Outcome: | The proposed method improves source coverage on the DiverseSumm benchmark. |
Bilingual Zero-Shot Stance Detection (2025.acl-long)
Copied to clipboard
| Challenge: | a study focuses on noun-phrase and claim targets within bilingual ZSSD scenarios . a dataset focusing on claim targets with a low occurrence of shared words is also explored . |
| Approach: | They use a bilingual bilingual ZSSD dataset to investigate the use of zero-shot stance detection. |
| Outcome: | The proposed dataset is the first to examine this difficult setting in bilingual ZSSD . it focuses on noun-phrase and claim targets within in-domain and out-of-domain bilingual scenarios . |
GrammaMT: Improving Machine Translation with Grammar-Informed In-Context Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Experiments show that GrammaMT enhances translation performance on open-source instruction-tuned LLMs for various low- to high-resource languages across three benchmarks: (1) largest corpus, (2) challenging 2023 SIGMORPHON Shared Task data, (3) even in an out-of-domain setting with FLORES. |
| Approach: | They propose a grammatically-aware prompting approach that uses Interlinear Glossed Text . they propose gloss-shot, chain-gloss and model-glooss prompting strategies that are training-free . |
| Outcome: | Experiments show that GrammaMT improves translation performance on open-source instruction-tuned LLMs for various low- to high-resource languages across three benchmarks. |
Theorem Prover as a Judge for Synthetic Data Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies show that large language models are increasingly capable of tackling mathematical problems. |
| Approach: | They propose an approach that iteratively refines theorem prover formalisation to mitigate errors. |
| Outcome: | The proposed method increases execution rate on the Lean prover from 60% to 87%, while human annotation is replaced with theorem prover feedback. |
Measuring the Effect of Transcription Noise on Downstream Language Understanding Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | a growing number of recorded human speech is recorded for automated processing, resulting in errors in the transcripts . a configurable framework is proposed to analyze transcript noise impact across noise levels and transcript-cleaning techniques. |
| Approach: | They propose a configurable framework for assessing task models in diverse noisy settings . framework facilitates investigation of task model behavior, which can support effective SLU solutions. |
| Outcome: | The proposed framework can analyze model behavior in various noise levels and transcript-cleaning techniques. |
Assessing Reliability and Political Bias In LLMs’ Judgements of Formal and Material Inferences With Partisan Conclusions (2025.acl-long)
Copied to clipboard
| Challenge: | This paper examines the ability of LLMs to correctly label simple inferences with partisan conclusions. |
| Approach: | They develop a dataset with formal and material inferences with conclusions that favor either the political left or the political right. |
| Outcome: | The proposed models show that they are unreliable and political bias persists throughout the English and German datasets. |
PARME: Parallel Corpora for Low-Resourced Middle Eastern Languages (2025.acl-long)
Copied to clipboard
Sina Ahmadi, Rico Sennrich, Erfan Karami, Ako Marani, Parviz Fekrazad, Gholamreza Akbarzadeh Baghban, Hanah Hadi, Semko Heidari, Mahîr Dogan, Pedram Asadi, Dashne Bashir, Mohammad Amin Ghodrati, Kourosh Amini, Zeynab Ashourinezhad, Mana Baladi, Farshid Ezzati, Alireza Ghasemifar, Daryoush Hosseinpour, Behrooz Abbaszadeh, Amin Hassanpour, Bahaddin Jalal Hamaamin, Saya Kamal Hama, Ardeshir Mousavi, Sarko Nazir Hussein, Isar Nejadgholi, Mehmet Ölmez, Horam Osmanpour, Rashid Roshan Ramezani, Aryan Sediq Aziz, Ali Salehi, Mohammadreza Yadegari, Kewyar Yadegari, Sedighe Zamani Roodsari
| Challenge: | UNESCO has identified 60 varieties of Middle Eastern languages as underrepresented . a limited availability of language technology perpetuates a cycle of digital exclusion . |
| Approach: | They develop a parallel corpora for eight severely under-resourced varieties in the region . they evaluate machine translation capabilities through zero-shot approaches and fine-tuning experiments . |
| Outcome: | The proposed model aims to improve the processing of the eight under-resourced languages in the Middle East. |
METAL: A Multi-Agent Framework for Chart Generation with Test-Time Scaling (2025.acl-long)
Copied to clipboard
| Challenge: | Chart generation requires strong visual design skills and precise coding capabilities that embed the desired visual properties into code. |
| Approach: | They propose a vision-language model-based multi-agent framework for effective automatic chart generation. |
| Outcome: | The proposed framework achieves a 5.2% improvement in the F1 score over the current best chart generation task. |
ConLoan: A Contrastive Multilingual Dataset for Evaluating Loanwords (2025.acl-long)
Copied to clipboard
Sina Ahmadi, Micha David Hess, Elena Álvarez-Mellado, Alessia Battisti, Cui Ding, Anne Göhring, Yingqiang Gao, Zifan Jiang, Andrianos Michail, Peshmerge Morad, Joel Niklaus, Maria Christina Panagiotopoulou, Stefano Perrella, Juri Opitz, Anastassia Shaitarova, Rico Sennrich
| Challenge: | Lexical borrowing is a ubiquitous linguistic phenomenon influenced by geopolitical, societal, and technological factors. |
| Approach: | They propose a novel contrastive dataset comprising sentences with and without loanwords across 10 languages to examine how machine translation and language models process loanword . |
| Outcome: | The proposed dataset shows that state-of-the-art models prefer loanwords over native terms and exhibit varying performance across languages. |
A Theory of Response Sampling in LLMs: Part Descriptive and Part Prescriptive (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly utilized in autonomous decision-making, where they sample options from vast action spaces. |
| Approach: | They propose to use heuristics to sample LLMs to identify a prescriptive component and a descriptive component to represent a concept. |
| Outcome: | The proposed model is compared with human decision-making models in public health, economic trends and other real-world domains to show that it is biased. |
MEraser: An Effective Fingerprint Erasure Approach for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have raised critical concerns about model ownership and intellectual property protection. |
| Approach: | They propose a method for effectively removing backdoor-based fingerprints from LLMs . they propose deleting backdoor fingerprints using a transferable erasure mechanism . |
| Outcome: | The proposed method removes backdoor-based fingerprints while maintaining model performance. |
VISA: Retrieval Augmented Generation with Visual Source Attribution (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to retrieval-augmented generation primarily link generated content to document-level references, making it difficult for users to locate evidence among multiple content-rich retrieved documents. |
| Approach: | They propose a novel approach that combines answer generation with visual source attribution by leveraging large vision-language models to identify evidence and highlight exact regions that support the generated answers with bounding boxes in the retrieved document screenshots. |
| Outcome: | The proposed approach identifies evidence and highlights exact regions that support the generated answers with bounding boxes in the retrieved document screenshots. |
DRAMA: Diverse Augmentation from Large Language Models to Smaller Dense Retrievers (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have shown strong effectiveness and robustness when fine-tuned as dense retrievers. |
| Approach: | They propose a training framework that leverages pruned LLMs to train smaller generalizable dense retrievers. |
| Outcome: | The proposed training framework offers better multilingual and long-context capabilities than traditional encoder-based retrievers and achieves strong performance across multiple tasks and languages. |
Stochastic Chameleons: Irrelevant Context Hallucinations Reveal Class-Based (Mis)Generalization in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have shown that LLMs reproduce training artifacts, exploit spurious correlations, and fail when faced with distribution shifts. |
| Approach: | They examine irrelevant context hallucinations in which models integrate misleading contextual cues into their predictions. |
| Outcome: | The proposed model errors are reflected in the model's internal computations, and they are consistent with previous studies. |
MAPoRL: Multi-Agent Post-Co-Training for Collaborative Large Language Models with Reinforcement Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on prompting and developing workflows with frozen LLMs. |
| Approach: | They propose a multi-agentic framework for collaborative LLMs with reinforcement learning that leverages multi-gendered frameworks to enhance collaboration. |
| Outcome: | The proposed model improves collaboration performance across multiple datasets with generalization to unseen domains compared to existing models. |
Map&Make: Schema Guided Text to Table Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for text-to-table generation overlook what complex information to extract and how to infer it from text. |
| Approach: | They propose a method that decomposes text into atomic propositions to infer latent schemas. |
| Outcome: | The proposed method shows significant gains in accuracy and interpretability on three datasets. |
IRIS: Interpretable Retrieval-Augmented Classification for Long Interspersed Document Sequences (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models for document classification struggle with long-text processing due to quadratic computational complexity in the self-attention module. |
| Approach: | They propose a framework that utilizes retrieval to efficiently classify long documents . they use a quadratic attention matrix to capture dependencies between tokens in an input sequence . |
| Outcome: | The proposed framework excels in clinical note disease risk prediction tasks . it can process arbitrarily long documents without increasing computational cost and trainable on a single GPU. |
Symmetrical Visual Contrastive Optimization: Aligning Vision-Language Models with Minimal Contrastive Images (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have shown that Large Vision-Language Models (VLMs) tend to neglect image content and over-rely on language-model priors, resulting in errors in visually grounded tasks and hallucinations. |
| Approach: | They propose a novel finetuning objective that steers the model toward capturing important visual details and aligning them with corresponding text tokens. |
| Outcome: | The proposed method achieves up to 22% reduction in hallucinations and significant gains in vision-centric and general tasks while maintaining or improving the model's general abilities. |
Can we Retrieve Everything All at Once? ARM: An Alignment-Oriented LLM-based Retrieval Method (2025.acl-long)
Copied to clipboard
| Challenge: | Existing RAG solutions address the alignment problem in a limited manner . ARM explores relationships among data objects, enabling a retrieve-all-at-once solution for complex queries . |
| Approach: | Experimental results show that ARM improves alignment of open-domain questions with available data . ARM explores relationships among data objects, enabling a retrieve-all-at-once solution for complex queries. |
| Outcome: | Experimental results show that ARM outperforms existing RAG methods on complex open-domain questions. |
R2D2: Remembering, Replaying and Dynamic Decision Making with a Reflective Agentic Memory (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for web agents struggle with efficient navigation and action execution due to limited visibility and understanding of web structures. |
| Approach: | They propose a framework that integrates memory-enhanced navigation and reflective learning to improve web agents' performance. |
| Outcome: | The proposed framework shows significant improvements over existing methods, including 50% reduction in navigation errors and threefold increase in task completion rates. |
FairI Tales: Evaluation of Fairness in Indian Contexts with a Focus on Bias and Stereotypes (2025.acl-long)
Copied to clipboard
Janki Atul Nawale, Mohammed Safi Ur Rahman Khan, Janani D, Mansi Gupta, Danish Pruthi, Mitesh M Khapra
| Challenge: | Existing studies on fairness of LLMs are largely Western-focused, making them inadequate for culturally diverse countries such as India. |
| Approach: | They propose a benchmark to evaluate fairness of LLMs across 85 identity groups . they consult domain experts to curate over 1,800 socio-cultural topics . |
| Outcome: | The benchmark evaluates LLMs across 85 identities across 85 castes, religions, regions, and tribes. |
SpeechIQ: Speech-Agentic Intelligence Quotient Across Cognitive Levels in Voice Understanding by Large Language Models (2025.acl-long)
Copied to clipboard
Zhen Wan, Chao-Han Huck Yang, Yahan Yu, Jinchuan Tian, Sheng Li, Ke Hu, Zhehuai Chen, Shinji Watanabe, Fei Cheng, Chenhui Chu, Sadao Kurohashi
| Challenge: | SIQ quantifies voice understanding abilities and provides unified comparisons between cascaded methods and end-to-end models. |
| Approach: | They propose a human cognition-inspired evaluation pipeline for voice understanding large language models (LLM_Voice) that quantifies voice understanding abilities and provides unified comparisons between cascaded methods and end-to-end models. |
| Outcome: | The proposed framework quantifies voice understanding abilities and provides unified comparisons between cascaded methods and end-to-end models, identifies annotation errors in existing benchmarks, and detects hallucinations in LLM_Voice. |
Predicting Implicit Arguments in Procedural Video Instructions (2025.acl-long)
Copied to clipboard
| Challenge: | Prior SRL benchmarks often miss implicit arguments, leading to incomplete understanding. |
| Approach: | They propose a dataset that necessitates inferring implicit and explicit arguments from contextual information in multimodal cooking procedures. |
| Outcome: | The proposed dataset achieves a 17% relative improvement in F1-score for what-implicit and a 14.7% improvement for where/with-implicative semantic roles over GPT-4o. |
PIGuard: Prompt Injection Guardrail via Mitigating Overdefense for Free (2025.acl-long)
Copied to clipboard
| Challenge: | Prompt injection attacks pose a critical threat to large language models, enabling goal hijacking and data leakage. |
| Approach: | They propose a prompt guard model that incorporates a new training strategy to mitigate over-defense for free . PIGuard significantly reduces the bias on trigger words, enabling fine-grained evaluation . |
| Outcome: | The proposed model outperforms the existing model on diverse benchmarks by 30.4%. |
CLIPErase: Efficient Unlearning of Visual-Textual Associations in CLIP (2025.acl-long)
Copied to clipboard
| Challenge: | MU has gained significant attention as a means to remove the influence of specific data from a trained model without requiring full retraining. |
| Approach: | They propose a novel approach that disentangles and selectively forgets both visual and textual associations, ensuring that unlearning does not compromise model performance. |
| Outcome: | Experiments on CIFAR-100, Flickr30K, and Conceptual 12M show that CLIPErase effectively removes designated associations from multimodal samples in downstream tasks while preserving model performance on retain set. |
ViGiL3D: A Linguistically Diverse Dataset for 3D Visual Grounding (2025.acl-long)
Copied to clipboard
| Challenge: | 3D visual grounding models localize entities in a scene referred to by natural language text . recent studies focused on LLM-based scaling of 3DVG datasets, but these do not capture the full range of potential prompts which could be specified in the English language. |
| Approach: | They propose a framework for linguistically analyzing 3DVG prompts and introduce a diagnostic dataset for evaluating 3D visual grounding methods against a diverse set of language patterns. |
| Outcome: | The proposed framework scales up and tests against a representative set of prompts in the english language. |
The time scale of redundancy between prosody and linguistic context (2025.acl-long)
Copied to clipboard
Tamar I Regev, Chiebuka Ohams, Shaylee Xie, Lukas Wolf, Evelina Fedorenko, Alex Warstadt, Ethan Wilcox, Tiago Pimentel
| Challenge: | Prior work has shown that the information carried by prosodic features is substantially redundant with that carried by the surrounding words. |
| Approach: | They examine the time scale of this relationship, studying how it varies with the length of past and future contexts. |
| Outcome: | The results show that prosody features show some redundancy with future words, but only with a short scale of 1-2 words, consistent with reports of incremental short-term planning in language production. |
Basic Reading Distillation (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models require high computational resources which limits their deployment in real-world applications. |
| Approach: | They propose to distill large language models into smaller language models by either knowledge distillation or task distillation. |
| Outcome: | The proposed model outperforms or performs comparable to over 20x bigger LLMs on language inference benchmarks and BIG-bench tasks. |
Quantized Can Still Be Calibrated: A Unified Framework to Calibration in Quantized Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to quantify uncertainty of large language models (LLMs) but their influence on uncertainty calibration remains unexplored. |
| Approach: | They propose an analytic method to estimate the upper bound of calibration error (UBCE) for quantized LLMs and propose a method to recover calibration errors through soft-prompt tuning. |
| Outcome: | The proposed method improves the calibration accuracy of quantized models on multiple datasets and LLMs. |
A Spatio-Temporal Point Process for Fine-Grained Modeling of Reading Behavior (2025.acl-long)
Copied to clipboard
| Challenge: | Standard models that focus on fixation durations ignore spatial dynamics of reading . authors propose a model that captures how long fixations last, where they land and when . |
| Approach: | They propose a generative model that captures how long fixations last and where they land and when they occur. |
| Outcome: | The proposed model exhibits higher likelihood on held-out reading data than baselines. |
More is not always better? Enhancing Many-Shot In-Context Learning with Differentiated and Reweighting Objectives (2025.acl-long)
Copied to clipboard
Xiaoqing Zhang, Ang Lv, Yuhan Liu, Flood Sung, Wei Liu, Jian Luan, Shuo Shang, Xiuying Chen, Rui Yan
| Challenge: | Large language models excel at few-shot in-context learning but performance plateaus as ICL demonstrations increase from a few to many. |
| Approach: | They propose a novel optimization method that optimizes the negative log-likelihood objective and reweights the model to achieve many-shot performance. |
| Outcome: | The proposed method achieves significant performance improvements across a large-scale dataset. |
Astute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have not linked the behavior of retrieval augmented generation (RAG) with imperfect retrieval, including irrelevant, misleading, or even malicious information. |
| Approach: | They propose an approach that integrates external knowledge with source-awareness to overcome imperfect retrieval errors in RAG. |
| Outcome: | The proposed approach is superior to previous robustness-enhanced approaches under the worst-case scenario. |
SubLIME: Subset Selection via Rank Correlation Prediction for Data-Efficient LLM Evaluation (2025.acl-long)
Copied to clipboard
Gayathri Saranathan, Cong Xu, Mahammad Parwez Alam, Tarun Kumar, Martin Foltin, Soon Yee Wong, Suparna Bhattacharya
| Challenge: | Large language models and datasets have made benchmark evaluations computationally prohibitive. |
| Approach: | They propose a framework that reduces evaluation costs by 80% to 99% while preserving ranking fidelity. |
| Outcome: | The proposed evaluation reduces evaluation costs by 80% to 99% while preserving ranking fidelity. |
M³GQA: A Multi-Entity Multi-Hop Multi-Setting Graph Question Answering Benchmark (2025.acl-long)
Copied to clipboard
| Challenge: | GraphRAG systems have achieved remarkable progress in enhancing performance and reliability of large language models. |
| Approach: | They propose a GraphRAG benchmark focusing on multi-entity queries with six settings for comprehensive evaluation. |
| Outcome: | The proposed method can construct diverse data with semantically correct ground-truth reasoning paths. |
LSSF: Safety Alignment for Large Language Models through Low-Rank Safety Subspace Fusion (2025.acl-long)
Copied to clipboard
| Challenge: | Existing safety alignment methods rely on fine-tuning, which inadvertently leads to the increased complexity and computational resources required. |
| Approach: | They propose a safety re-alignment framework with Low-Rank Safety Subspace Fusison that exploits low-rank safety characteristics of LLMs by constructing a low-ranked projection matrix to extract the principal components of safety vectors. |
| Outcome: | The proposed method exploits low-rank safety subspace of the LLMs and is stable during fine-tuning process and is isolated from the model’s general capabilities. |
ETF: An Entity Tracing Framework for Hallucination Detection in Code Summaries (2025.acl-long)
Copied to clipboard
Kishan Maharaj, Vitobha Munigala, Srikanth G. Tamilselvam, Prince Kumar, Sayandeep Sen, Palani Kodeswaran, Abhijit Mishra, Pushpak Bhattacharyya
| Challenge: | Recent advances in large language models have significantly enhanced their ability to understand both natural language and code, but are prone to hallucinations. |
| Approach: | They propose a first-of-its-kind dataset, CodeSumEval, with 10K samples, curated specifically for hallucination detection in code summarisation. |
| Outcome: | The proposed framework has a 73% F1 score and is curated specifically for detection of hallucinations in code summarisation. |
Meta-Tool: Unleash Open-World Function Calling Capabilities of General-Purpose Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models struggle with addressing diverse user inquiries in open-world tasks. |
| Approach: | They propose a plug-and-play tool retrieval system for LLMs to access external tool library and use retrieved tools to solve user's problem. |
| Outcome: | The proposed model improves on a finetuned version of LLaMA-3.1 and 2,800 dialogues and 7,361 tools spanning ten distinct test categories. |
Benchmarking and Improving Large Vision-Language Models for Fundamental Visual Graph Understanding and Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for enhancing understanding and reasoning abilities in graphbased tasks focus on specific graph types or tasks, posing challenges in designing versatile systems suitable for various tasks and graphs across diverse domains. |
| Approach: | They propose a structure-aware fine-tuning framework to enhance LVLMs with structure learning abilities through three self-supervised learning tasks. |
| Outcome: | Extensive evaluations on 14 LVLMs reveal that LVLs are weak in basic graph understanding and reasoning tasks, particularly those concerning relational or structurally complex information. |
ISR: Self-Refining Referring Expressions for Entity Grounding (2025.acl-long)
Copied to clipboard
| Challenge: | Entity grounding is a crucial task in the construction of multimodal knowledge graphs. |
| Approach: | They propose a novel scheme to enhance the multimodal large language model's capability to generate high quality REs for the given entities as explicit contextual clues. |
| Outcome: | The proposed method surpasses other methods in entity grounding, highlighting its effectiveness, robustness and potential for broader applications. |
Activating Distributed Visual Region within LLMs for Efficient and Effective Vision-Language Training and Inference (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Large Vision-Language Models (LVLMs) learn visual capacity through visual instruction tuning. |
| Approach: | They propose a method for LVLMs to be trained by selective layers tuning . they propose removing non-critical layers outside the visual region . |
| Outcome: | The proposed approach preserves nearly 99% of visual performance and improves textual task results while reducing training time. |
CCHall: A Novel Benchmark for Joint Cross-Lingual and Cross-Modal Hallucinations Detection in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies on hallucinations in large language models are limited to a single scenario, either cross-lingual or cross-modal. |
| Approach: | They propose a joint Cross-lingual and Cross-modal hallucinations benchmark to fill this gap . they incorporate cross-lingual, cross-modal scenarios to assess hallucinic capabilities . |
| Outcome: | The proposed benchmark incorporates both cross-lingual and cross-modal hallucination scenarios to assess the cross-linguistic and crossmodal capabilities of LLMs. |
TestNUC: Enhancing Test-Time Computing Approaches and Scaling through Neighboring Unlabeled Data Consistency (2025.acl-long)
Copied to clipboard
Henry Peng Zou, Zhengyao Gu, Yue Zhou, Yankai Chen, Weizhi Zhang, Liancheng Fang, Yibo Wang, Yangning Li, Kay Liu, Philip S. Yu
| Challenge: | Test-time computing approaches that leverage additional computational resources during inference have been proven effective in enhancing large language model performance. |
| Approach: | They propose a linearly scaling approach that leverages local consistency of neighboring unlabeled data to improve test-time predictions. |
| Outcome: | The proposed approach outperforms baseline methods such as prompting and self-consistency across eight datasets and performs robustly across embedding models. |
The Esethu Framework: Reimagining Sustainable Dataset Governance and Curation for Low-Resource Languages (2025.acl-long)
Copied to clipboard
Jenalea Rajab, Anuoluwapo Aremu, Everlyn Asiko Chimoto, Dale Dunbar, Graham Morrissey, Fadel Thior, Luandrie Potgieter, Jessica Ojo, Atnafu Lambebo Tonja, Wilhelmina NdapewaOnyothi Nekoto, Pelonomi Moiloa, Jade Abbott, Vukosi Marivate, Benjamin Rosman
| Challenge: | Esethu Framework is a community-centric data license that empowers local communities and ensures equitable benefit-sharing from their linguistic resource. |
| Approach: | They propose a community-centric data license to empower local communities and ensure equitable benefit-sharing from their linguistic resource. |
| Outcome: | The proposed dataset contains read speech from native isiXhosa speakers enriched with demographic and linguistic metadata. |
Theoretical Analysis of Hierarchical Language Recognition and Generation by Transformers without Positional Encoding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that Transformers can recognize hierarchical structures without a specific positional encoding. |
| Approach: | They show that Transformers can generate hierarchical languages without a positional encoding . they also suggest that explicit positional encoders might have a detrimental effect on generalization . |
| Outcome: | The proposed model can generate hierarchical languages with respect to model size without encoding . |
Less is More: Explainable and Efficient ICD Code Prediction with Clinical Entities (2025.acl-long)
Copied to clipboard
| Challenge: | Clinical coding is labor-intensive and prone to delays, leading to global backlogs. |
| Approach: | They propose an approach that combines Named Entity Recognition (NER) and Assertion Classification (AC) to filter for clinically important content before supervised code prediction. |
| Outcome: | The proposed approach reduces training time by over half on a standard evaluation dataset compared to current methods . it uses Named Entity Recognition (NER) and Assertion Classification (AC) to filter for clinically important content before supervised code prediction. |
Benchmarking LLMs and LLM-based Agents in Practical Vulnerability Detection for Code Repositories (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown promise in software vulnerability detection, especially on function-level benchmarks like Devign and BigVul. |
| Approach: | They propose a JIT vulnerability detection benchmark linking each function to its vulnerability-introducing and fixing commits. |
| Outcome: | The proposed JIT vulnerability detection benchmark enables comprehensive evaluation of detection capabilities. |
Multi-Modality Expansion and Retention for LLMs through Parameter Merging and Decoupling (2025.acl-long)
Copied to clipboard
Junlin Li, Guodong Du, Jing Li, Sim Kuan Goh, Wenya Wang, Yequan Wang, Fangming Liu, Ho-Kin Tang, Saleh Alharbi, Daojing He, Min Zhang
| Challenge: | Large Language Models (LLMs) are a cornerstone in artificial intelligence due to their exceptional performance. |
| Approach: | They propose a training-free approach that integrates existing MLLMs for effective multimodal expansion while retaining their original performance. |
| Outcome: | The proposed approach can expand LLMs' multimodal capabilities while retaining original performance. |
Serial Lifelong Editing via Mixture of Knowledge Experts (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Lifelong Knowledge Editing methods struggle to overwrite outdated knowledge with the latest one. |
| Approach: | They propose a new Mixture-of-Knowledge-Experts scheme with an ARM . ARM ensures that each update completely overwrites old information with the latest one . Experimental results show that ARM performs favorably against SOTA knowledge editing methods . |
| Outcome: | The proposed scheme overwrites old knowledge with the latest data on a benchmark . it performs favorably against existing knowledge editing methods on the same concept . |
A Survey on Efficient Large Language Model Training: From Data-centric Perspectives (2025.acl-long)
Copied to clipboard
Junyu Luo, Bohan Wu, Xiao Luo, Zhiping Xiao, Yiqiao Jin, Rong-Cheng Tu, Nan Yin, Yifan Wang, Jingyang Yuan, Wei Ju, Ming Zhang
| Challenge: | achieving data-efficient post-training of Large Language Models is a key research question. |
| Approach: | They propose a taxonomy of data-efficient LLM post-training methods from a data-centric perspective. |
| Outcome: | The proposed methods cover data selection, data quality enhancement, synthetic data generation, data distillation and compression, and self-evolving data ecosystems. |
IMOL: Incomplete-Modality-Tolerant Learning for Multi-Domain Fake News Video Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fake news video detection focus on a specific domain and assume multiple modalities. |
| Approach: | They propose an incomplete-modality-tolerant learning framework for fake news video detection . they use cross-modal consistency to reconstruct missing modalities and transferable knowledge through cross-sample reasoning . |
| Outcome: | The proposed framework improves performance and robustness of multi-domain fake news video detection while generalizing to unseen domains under incomplete modality conditions. |
DDxTutor: Clinical Reasoning Tutoring System with Differential Diagnosis-Based Structured Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) have enabled various medical educational applications, but they often provide direct answers that could reduce students’ cognitive engagement and lead to fragmented learning. |
| Approach: | They propose a framework that follows differential diagnosis principles to decompose clinical reasoning into teachable components. |
| Outcome: | The proposed framework decomposes clinical reasoning into teachable components and generates structured teaching references and conducts diagnostic tutoring dialogues. |
SocialEval: Evaluating Social Intelligence of Large Language Models (2025.acl-long)
Copied to clipboard
Jinfeng Zhou, Yuxuan Chen, Yihan Shi, Xuanming Zhang, Leqi Lei, Yi Feng, Zexuan Xiong, Miao Yan, Xunzhi Wang, Yaru Cao, Jianing Yin, Shuai Wang, Quanyu Dai, Zhenhua Dong, Hongning Wang, Minlie Huang
| Challenge: | Existing work on LLMs does not address their social intelligence (SI) and their discrepancy with humans. |
| Approach: | They propose a script-based bilingual SI benchmark that integrates outcome-oriented goal achievement evaluation and process-oriented interpersonal ability evaluation by manually crafting narrative scripts. |
| Outcome: | The proposed model is based on a script-based bilingual evaluation paradigm that integrates outcome- and process-oriented evaluation by manually crafting narrative scripts. |
Hidden in Plain Sight: Evaluation of the Deception Detection Capabilities of LLMs in Multimodal Settings (2025.acl-long)
Copied to clipboard
| Challenge: | Detecting deception in an increasingly digital world is a critical and challenging task. |
| Approach: | They evaluate the performance of both open-source and proprietary LLMs on three datasets . they find that fine-tuned LLM achieve state-of-the-art performance on textual deception detection . |
| Outcome: | The proposed models achieve state-of-the-art on textual deception detection, whereas LMMs struggle to fully leverage multimodal cues. |
Analyzing and Mitigating Inconsistency in Discrete Speech Tokens for Neural Codec Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) have demonstrated significant strides in generating high-quality speech . discretizing speech by neural audio codecs often results in sequences that differ from text sequences . |
| Approach: | They quantitatively analyze the Discrete Representation Inconsistency phenomenon within popular audio tokenizers such as EnCodec. |
| Outcome: | The proposed method mitigates the DRI phenomenon within popular audio tokenizers such as EnCodec. |
PlanningArena: A Modular Benchmark for Multidimensional Evaluation of Planning and Tool Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have shown that LLMs can be significantly improved by integrating external tools. |
| Approach: | They propose a framework that integrates external tools into large language models to evaluate their ability to generate action plans. |
| Outcome: | The proposed framework evaluates the ability of large language models to generate action plans and generate action plan templates. |
FocusLLM: Precise Understanding of Long Context by Dynamic Condensing (2025.acl-long)
Copied to clipboard
Zhenyu Li, Yike Zhang, Tengyu Pan, Yutao Sun, Zhichao Duan, Junjie Fang, Rong Han, Zixuan Wang, Jianyong Wang
| Challenge: | Existing context condensing methods cannot accurately understand the full context, as there is a considerable amount of information loss in the condensed process. |
| Approach: | They propose a framework to extend the fixed context length of any decoder-only LLM by distilling crucial information from long sequences. |
| Outcome: | The proposed framework extends the fixed context length of any decoder-only LLM, allowing it to focus on relevant information from very long sequences. |
Negative Matters: Multi-Granularity Hard-Negative Synthesis and Anchor-Token-Aware Pooling for Enhanced Text Embeddings (2025.acl-long)
Copied to clipboard
| Challenge: | Text embedding models are used for various natural language processing tasks such as sentiment analysis, text clustering, and content-based information retrieval. |
| Approach: | They propose a synthesis framework that leverages large language models to generate diverse negative samples with varying levels of similarity with the query. |
| Outcome: | The proposed framework achieves state-of-the-art performance surpassing existing synthesis strategies with synthetic data and when combined with public retrieval datasets. |
GPT-4 as a Homework Tutor Can Improve Student Engagement and Learning Outcomes (2025.acl-long)
Copied to clipboard
| Challenge: | a recent study has shown that homework is never graded or is done superficially. |
| Approach: | They propose a prompting strategy that enables GPT-4 to conduct interactive homework sessions for high school students learning English as a second language. |
| Outcome: | The proposed solution improves homework in high school students learning English as a second language with minimal effort in content preparation, one of the key challenges of alternative methods. |
Diffusion Models Through a Global Lens: Are They Culturally Inclusive? (2025.acl-long)
Copied to clipboard
Zahra Bayramli, Ayhan Suleymanzade, Na Min An, Huzama Ahmad, Eunsu Kim, Junyeong Park, James Thorne, Alice Oh
| Challenge: | Text-to-image diffusion models have produced compelling, detailed images from text prompts, but their ability to accurately represent cultural nuances remains an open question. |
| Approach: | They propose a benchmark to evaluate whether diffusion models can generate culturally specific images spanning ten countries. |
| Outcome: | The proposed model fails to generate culturally specific images spanning ten countries . it shows significant disparities in cultural relevance, description fidelity, and realism compared to real-world reference images. |
Efficient Safety Alignment of Large Language Models via Preference Re-ranking and Representation-based Reward Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Existing safety alignment methods for Large Language Models (LLMs) face the distribution shift issue, which requires significant computational resources. |
| Approach: | They propose a framework that leverages the model’s intrinsic safety judgment capability to extract reward signals, which are then used to calculate label confidence for preference reordering. |
| Outcome: | The proposed framework improves safety performance while avoiding 300x computational overheads. |
English-based acoustic models perform well in the forced alignment of two English-based Pacific Creoles (2025.acl-long)
Copied to clipboard
Sam Passmore, Lila San Roque, Kirsty Gillespie, Saurabh Nath, Kira Davey, Keira Mullan, Tim Cawley, Jennifer Biggs, Rosey Billington, Bethwyn Evans, Nick Thieberger, Danielle Barth
| Challenge: | Currently, European languages dominate phonetic research . forced alignment can accelerate the study of sociophonetic variation in minority languages . |
| Approach: | They propose to use English and custom-made acoustic models to study the alignment of vowels in two Pacific Creoles, Tok Pisin and Bislama. |
| Outcome: | The proposed models perform acceptablely well in English and humans in vowel environments described as ‘Highly Reliable’. |
Subtle Errors in Reasoning: Preference Learning via Error-injected Self-editing (2025.acl-long)
Copied to clipboard
Kaishuai Xu, Tiezheng Yu, Wenjun Hou, Yi Cheng, Chak Tou Leong, Liangyou Li, Xin Jiang, Lifeng Shang, Qun Liu, Wenjie Li
| Challenge: | Existing studies to improve mathematical ability typically involve applying preference learning to step-wise solution pairs, but they overlook critical subtle errors. |
| Approach: | They propose a preference learning framework that injects predefined subtle errors into pivotal tokens to construct hard pairs for error mitigation. |
| Outcome: | Extensive experiments show that the proposed framework improves on Qwen2-7B-Instruct and MATH with 4.5K training samples. |
Truth Knows No Language: Evaluating Truthfulness Beyond English (2025.acl-long)
Copied to clipboard
Blanca Calvo Figueras, Eneko Sagarzazu, Julen Etxaniz, Jeremy Barnes, Pablo Gamallo, Iria de-Dios-Flores, Rodrigo Agerri
| Challenge: | a new benchmark evaluates the truthfulness of large language models (LLMs) based on imitative falsehoods. |
| Approach: | They propose a professionally translated extension of the TruthfulQA benchmark . it evaluates truthfulness in Basque, Catalan, Galician, and Spanish . |
| Outcome: | The proposed extension of the TruthfulQA benchmark evaluates truthfulness in Basque, Catalan, Galician, and Spanish. |
Revisiting Compositional Generalization Capability of Large Language Models Considering Instruction Following Ability (2025.acl-long)
Copied to clipboard
| Challenge: | generative large language models (LLMs) compose sentences that include all given concepts but must generate sentences that adhere to the specified order. |
| Approach: | They propose a benchmark to evaluate compositional generalization and instruction-following abilities of generative large language models (LLMs) based on ordered coverage, which allows simultaneous evaluation of both abilities. |
| Outcome: | The proposed benchmark evaluates compositional generalization and instruction-following abilities of LLMs. |
Batayan: A Filipino NLP benchmark for evaluating Large Language Models (2025.acl-long)
Copied to clipboard
Jann Railey Montalan, Jimson Paulo Layacan, David Demitri Africa, Richell Isaiah S. Flores, Michael T. Lopez Ii, Theresa Denise Magsajo, Anjanette Cayabyab, William Chandra Tjhi
| Challenge: | Recent advances in large language models (LLMs) have demonstrated remarkable capabilities on widely benchmarked high-resource languages. |
| Approach: | They propose a benchmark that systematically evaluates LLMs across three key natural language processing competencies: understanding, reasoning, and generation. |
| Outcome: | The proposed benchmark covers eight tasks covering Tagalog and code-switched Taglish utterances. |
HintsOfTruth: A Multimodal Checkworthiness Detection Dataset with Real and Synthetic Claims (2025.acl-long)
Copied to clipboard
| Challenge: | Identifying checkworthy claims is the first step, but detection methods struggle with content that is (1) multimodal, (2) from diverse domains, and (3) synthetic. |
| Approach: | They propose a dataset for multimodal checkworthiness detection with 27K real-world and synthetic image/claim pairs. |
| Outcome: | The proposed dataset compares lightweight text-based encoders to multimodal models but only focus on claim-like content. |
CityNavAgent: Aerial Vision-and-Language Navigation with Hierarchical Semantic Planning and Global Memory (2025.acl-long)
Copied to clipboard
Weichen Zhang, Chen Gao, Shiquan Yu, Ruiying Peng, Baining Zhao, Qian Zhang, Jinqiang Cui, Xinlei Chen, Yong Li
| Challenge: | Existing ground VLN agents struggle in aerial VLLN due to the lack of predefined navigation graphs and the exponentially expanding action space in long-horizon exploration. |
| Approach: | They propose a large language model-empowered aerial VLN agent that decomposes the long-horizon task into sub-goals with different semantic levels. |
| Outcome: | The proposed method achieves state-of-the-art performance with significant improvement in continuous city environments. |
It’s Not a Walk in the Park! Challenges of Idiom Translation in Speech-to-text Systems (2025.acl-long)
Copied to clipboard
| Challenge: | idioms are defined as words with a figurative meaning not deducible from their individual components. |
| Approach: | They compare idiom translation as compared to conventional news translation in two languages . they compare MT and SLT systems with MT, Large Language Models and cascaded alternatives . |
| Outcome: | The proposed systems show better handling of idioms than standard news translation systems. |
PolyNarrative: A Multilingual, Multilabel, Multi-domain Dataset for Narrative Extraction from News Articles (2025.acl-long)
Copied to clipboard
Nikolaos Nikolaidis, Nicolas Stefanovitch, Purificação Silvano, Dimitar Iliyanov Dimitrov, Roman Yangarber, Nuno Guimarães, Elisa Sartori, Ion Androutsopoulos, Preslav Nakov, Giovanni Da San Martino, Jakub Piskorski
| Challenge: | a new dataset of news articles annotated for narratives provides a framework for narrative detection . recurring narratives can propagate with very high velocity across audiences, languages and countries . |
| Approach: | They propose a multilingual dataset annotated for narratives using two-level taxonomies . they define narrative as a recurring, repetitive, overt or implicit claim that promotes a specific interpretation or viewpoint on an ongoing topic . |
| Outcome: | The proposed dataset will foster research in narrative detection and enable new research directions . the authors identify multiple narratives in the same article, and the results are published online . |
A Parameter-Efficient and Fine-Grained Prompt Learning for Vision-Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Current vision-language models extract semantic information from large-scale cross-modal associations, limiting performance and efficiency. |
| Approach: | They propose a detail-oriented prompt learning method to implement fine-grained multi-modal semantic alignment with merely 0.25M trainable parameters. |
| Outcome: | The proposed method implements fine-grained multi-modal semantic alignment with merely 0.25M trainable parameters. |
Persona Dynamics: Unveiling the Impact of Persona Traits on Agents in Text-Based Games (2025.acl-long)
Copied to clipboard
| Challenge: | Text-based interactive environments have long presented formidable challenges for AI. |
| Approach: | They propose a method for projecting human personality traits onto agents to guide their behavior and integrate them into their policy-learning pipelines. |
| Outcome: | The proposed method induces personality in a text-based game agent by integrating personality profiles directly into the agent's policy-learning pipeline. |
SeedBench: A Multi-task Benchmark for Evaluating Large Language Models in Seed Science (2025.acl-long)
Copied to clipboard
Jie Ying, Zihong Chen, Zhefan Wang, Wanli Jiang, Chenyang Wang, Zhonghang Yuan, Haoyang Su, Huanjun Kong, Fan Yang, Nanqing Dong
| Challenge: | Seed science is essential for modern agriculture, but its application in seed science remains limited due to a shortage of experts and limited availability of online resources. |
| Approach: | They evaluate 26 leading large language models and compare them against a set of benchmarks . they find that there is a gap between the power of LLMs and real-world seed science problems . |
| Outcome: | The new seed benchmark highlights the gap between the power of large language models and real-world seed science problems. |
-Stance: A Large-Scale Real World Dataset of Stances in Legal Argumentation (2025.acl-long)
Copied to clipboard
| Challenge: | Current tools for legal argument reasoning do not support this task. |
| Approach: | They propose to use a large-scale dataset to facilitate work on the legal argument stance classification task by evaluating whether a case summary strengthens or weakens a legal argument. |
| Outcome: | The proposed dataset is used to facilitate work on the legal argument stance classification task, which involves assessing whether a case summary strengthens or weakens a legal argument (polarity) and to what extent (intensity). |
Re3Syn: A Dependency-Based Data Synthesis Framework for Long-Context Post-training (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for constructing long-context data by concatenating short documents have overlooked a crucial characteristic of long-constituency data quality, semantic dependency. |
| Approach: | They propose a framework called Retrieval, Dependency Recognition, and Reorder for data synthesis which leverages semantic similarity to retrieve relevant documents and form several batches. |
| Outcome: | The proposed framework leverages semantic similarity to retrieve relevant documents and form several batches. |
Enabling Chatbots with Eyes and Ears: An Immersive Multimodal Conversation System for Dynamic Interactions (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodality has been explored in multi-party and multi-session conversations, but task-specific constraints have hindered its seamless integration into dynamic, natural conversations. |
| Approach: | They propose a multimodal conversation dataset and a model with multimodal memory retrieval to equip chatbots with "eyes and ears" they aim to integrate multimodality into chatbot interactions by integrating visual and auditory inputs into the chatbot. |
| Outcome: | The proposed model demonstrates the ability to engage in long-term conversations with multiple speakers in complex, real-world-like settings, effectively processing visual and auditory inputs to understand and respond appropriately. |
Multimodal Coreference Resolution for Chinese Social Media Dialogues: Dataset and Benchmark Approach (2025.acl-long)
Copied to clipboard
| Challenge: | Multimodal coreference resolution (MCR) aims to identify mentions referring to the same entity across different modalities, such as text and visuals. |
| Approach: | They propose a Chinese multimodal coreference dataset based on Douyin short-video platform to help researchers understand multimodal content. |
| Outcome: | The proposed dataset pairs short videos with corresponding textual dialogues from user comments and includes manually annotated coreference clusters for person mentions in the text and the coreferential person head regions in the corresponding video frames. |
TACLR: A Scalable and Efficient Retrieval-based Method for Industrial Product Attribute Value Identification (2025.acl-long)
Copied to clipboard
Yindu Su, Huike Zou, Lin Sun, Ting Zhang, Haiyang Yang, Chen Li Yu, David Lo, Qingheng Zhang, Shuguang Han, Jufeng Chen
| Challenge: | Existing methods for product attribute value identification face critical challenges . seller-provided attribute values are often incomplete or inaccurate . |
| Approach: | They propose a retrieval-based method that uses taxonomy-aware contrastive learning . they use product profiles and candidate values to encode and retrieve attributes based on similarity . |
| Outcome: | The proposed method is based on a taxonomy-aware, hard negative sampling and adaptive inference with dynamic thresholds. |
Theory of Mind in Large Language Models: Assessment and Enhancement (2025.acl-long)
Copied to clipboard
| Challenge: | Theory of Mind (ToM) is a cornerstone of human social intelligence . Large Language Models (LLMs) are increasingly integrated into daily life . |
| Approach: | They analyze evaluation benchmarks and enhancement strategies to evaluate LLMs' ToM capabilities. |
| Outcome: | The proposed and widely used story-based benchmarks and enhancement strategies are used to evaluate LLMs' ToM capabilities. |
Completing A Systematic Review in Hours instead of Months with Interactive AI Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Systematic reviews (SRs) are vital for evidence-based practice in high stakes disciplines, such as healthcare. |
| Approach: | They propose a human-centered interactive AI agent powered by large language models that partitions a large literature corpus based on semantics and employs . |
| Outcome: | InsightAgent improves quality of synthesized SRs by 27.2%, reaching 79.7% of human-written quality. |
CMHKF: Cross-Modality Heterogeneous Knowledge Fusion for Weakly Supervised Video Anomaly Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods focus mainly on visual modalities, neglecting rich multi-modality information. |
| Approach: | They propose a framework that integrates cross-modality knowledge from video, audio and text to improve anomaly detection and localization. |
| Outcome: | The proposed framework improves detection and localization of anomalies using video-level labels. |
CLaSp: In-Context Layer Skip for Self-Speculative Decoding (2025.acl-long)
Copied to clipboard
Longze Chen, Renke Shan, Huiming Wang, Lu Wang, Ziqiang Liu, Run Luo, Jiawei Wang, Hamid Alinejad-Rokny, Min Yang
| Challenge: | Existing methods for drafting Large Language Models require additional modules to be trained, which can be challenging to implement and ensure compatibility across various LLMs. |
| Approach: | They propose an in-context layer-skipping strategy for self-speculative decoding that uses a plug-and-play mechanism to skip intermediate layers of the verify model to construct a compressed draft model. |
| Outcome: | The proposed method achieves a speedup of 1.3 1.7 on LLaMA3 series models without altering the original distribution of the generated text. |
Teaching Text Agents to Learn Sequential Decision Making from Failure (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve text-based reinforcement-learning agents' performance contain failed actions that reinforce incorrect behaviors and reduce task success rates. |
| Approach: | They propose a failed action-aware objective that suppresses negative impact of failed actions . they propose 'failed action-based' perturbation method that leverages unsuccessful trajectories to construct new successful ones . |
| Outcome: | The proposed method outperforms baselines and generalizes across environments. |
The Harmonic Structure of Information Contours (2025.acl-long)
Copied to clipboard
Eleftheria Tsipidi, Samuel Kiegeland, Franz Nowak, Tianyang Xu, Ethan Wilcox, Alex Warstadt, Ryan Cotterell, Mario Giulianelli
| Challenge: | Language typically does not maintain a uniform information rate, but it fluctuates around a global average . a new study suggests periodicity may be a factor in information rate oscillations . |
| Approach: | They propose a hypothesis that language does not maintain a uniform information rate . they apply harmonic regression and introduce a new extension to detect periodicity . |
| Outcome: | The proposed method reveals that language oscillates at periodic intervals across frequencies . it also offers a framework for uncovering structural pressures at various levels of linguistic granularity. |
REAL-MM-RAG: A Real-World Multi-Modal Retrieval Benchmark (2025.acl-long)
Copied to clipboard
Navve Wasserman, Roi Pony, Oshri Naparstek, Adi Raz Goldfarb, Eli Schwartz, Udi Barzelay, Leonid Karlinsky
| Challenge: | Existing benchmarks do not fully capture real-world retrieval challenges . existing benchmarks lack a complete understanding of how models perform in realistic setups . |
| Approach: | They propose an automatic benchmark to address four key properties essential for real-world retrieval: (i) multi-modal documents, (ii) enhanced difficulty, ( (iv) Realistic-RAG queries and (v) accurate labeling. |
| Outcome: | The proposed model reveals significant model weaknesses, particularly in handling table-heavy documents and robustness to query rephrasing. |
Only a Little to the Left: A Theory-grounded Measure of Political Bias in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Political biases in language models can affect performance in many applications . political biased models are often left-leaning, but are generally more left- leaning for instruction-tuned models . |
| Approach: | They propose to use the Political Compass Test to measure political bias in language models . they use survey-based evaluation tools to test prompts and classify their political stances . |
| Outcome: | The proposed model is based on the Political Compass Test, but is not scientifically valid. |
LongSafety: Evaluating Long-Context Safety of Large Language Models (2025.acl-long)
Copied to clipboard
Yida Lu, Jiale Cheng, Zhexin Zhang, Shiyao Cui, Cunxiang Wang, Xiaotao Gu, Yuxiao Dong, Jie Tang, Hongning Wang, Minlie Huang
| Challenge: | Large Language Models (LLMs) have demonstrated remarkable capabilities in understanding and generating long sequences. |
| Approach: | They propose a benchmark to evaluate LLM safety in open-ended long-context tasks . they find that relevant context and extended input sequences can exacerbate safety risks . |
| Outcome: | The proposed benchmark identifies significant safety vulnerabilities in 16 LLMs . strong safety performance in short-context scenarios does not correlate with safety in long-contact tasks . |
Exploiting Contextual Knowledge in LLMs through 𝒱-usable Information based Layer Enhancement (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to enhance the context-faithfulness of Large Language Models (LLMs) ignore the fundamental mechanism of how contextual information is processed within LLMs’ internal states. |
| Approach: | They propose a method that enhances the utilization of contextual knowledge within LLMs’ internal representations by employing V-usable information analysis. |
| Outcome: | The proposed method improves context-faithfulness generation in Question-Answering tasks, particularly in scenarios involving unknown or conflicting contextual knowledge. |
Unintended Harms of Value-Aligned LLMs: Psychological and Empirical Insights (2025.acl-long)
Copied to clipboard
| Challenge: | Value-aligned LLMs are more prone to harmful behavior than fine-tuned models . value-aligned models generate text according to the aligned values, which can amplify harmful outcomes. |
| Approach: | They propose to use in-context alignment methods to enhance the safety of value-aligned LLMs. |
| Outcome: | The proposed methods improve value alignment and safety, the authors say . value-aligned models are more prone to harmful behavior than fine-tuned models . |
Maximal Matching Matters: Preventing Representation Collapse for Robust Cross-Modal Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to cross-modal image-text retrieval struggle with nuanced cross-modal relationships. |
| Approach: | They propose a set-based approach that represents each sample with multiple embeddings to capture nuanced and diverse relationships. |
| Outcome: | The proposed method achieves state-of-the-art performance on MS-COCO and Flickr30k without external data. |
The Noisy Path from Source to Citation: Measuring How Scholars Engage with Past Research (2025.acl-long)
Copied to clipboard
| Challenge: | Academic citations are widely used for evaluating research and tracing knowledge flows. |
| Approach: | They propose a computational pipeline to quantify citation fidelity at the sentence level by identifying citations in citing papers and corresponding claims in cited papers. |
| Outcome: | The proposed pipeline identifies citations in citing papers and the corresponding claims in cited papers and applies supervised models to measure fidelity at the sentence level. |
MAPLE: Enhancing Review Generation with Multi-Aspect Prompt LEarning in Explainable Recommendation (2025.acl-long)
Copied to clipboard
Ching-Wen Yang, Zhi-Quan Feng, Ying-Jia Lin, Che Wei Chen, Kun-da Wu, Hao Xu, Yao Jui-Feng, Hung-Yu Kao
| Challenge: | Existing models that generate generic aspects do not provide personalized informative recommendations. |
| Approach: | They propose a model that integrates aspect category as another input dimension to facilitate memorizing fine-grained aspect terms. |
| Outcome: | The proposed model outperforms baseline model on restaurant review datasets in the restaurant domain. |
Separating Tongue from Thought: Activation Patching Reveals Language-Agnostic Concept Representations in Transformers (2025.acl-long)
Copied to clipboard
| Challenge: | A central question in multilingual language modeling is whether large language models develop a universal concept representation, disentangled from specific languages. |
| Approach: | They analyze latent representations during a word-translation task in transformer-based LLMs and extract the residual stream of the last token of the word to be translated and insert the mean at the corresponding positions in the forward pass. |
| Outcome: | The proposed model can translate a word in multiple languages without changing the language and vice versa. |
Behavioural vs. Representational Systematicity in End-to-End Models: An Opinionated Survey (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks and models focus on systematicity of representations, but they focus on the systematicity in behaviour. |
| Approach: | They argue that systematicity is a desirable property in ML models as it enables strong generalization to novel contexts. |
| Outcome: | The proposed benchmarks and models focus on the systematicity of behaviour, while existing models focus primarily on language and vision. |
Dynamic Chunking and Selection for Reading Comprehension of Ultra-Long Context in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Current methods for improving large language models rely on splitting long contexts into fixed-length chunks, compromising accuracy. |
| Approach: | They propose a method for dynamically separating and selecting chunks of long context, facilitating a more streamlined input for LLMs. |
| Outcome: | The proposed approach outperforms baseline methods on single-hop and multi-hop question-answering benchmarks. |
DualRAG: A Dual-Process Approach to Integrate Reasoning and Retrieval for Multi-Hop Question Answering (2025.acl-long)
Copied to clipboard
Rong Cheng, Jinyi Liu, Yan Zheng, Fei Ni, Jiazhen Du, Hangyu Mao, Fuzheng Zhang, Bo Wang, Jianye Hao
| Challenge: | Existing approaches to multi-hop question answering struggle to identify and organize dynamic knowledge . et al., 2023; Liu e.t. al. 2023) suggest a dual-process framework for multi-step reasoning . |
| Approach: | They propose a synergistic dual-process framework that integrates reasoning and retrieval. |
| Outcome: | The proposed framework improves answer accuracy and coherence even in smaller-scale models. |
Deliberate Reasoning in Language Models as Structure-Aware Planning with an Accurate World Model (2025.acl-long)
Copied to clipboard
| Challenge: | Existing Chain-of-Thought (CoT) methods struggle with consistency and verification in complex reasoning tasks. |
| Approach: | They propose a framework that integrates structured knowledge representation with learned planning. |
| Outcome: | The proposed framework outperforms existing Chain-of-Thought (CoT) methods on math reasoning, logical reasoning, and coding tasks. |
Refining Salience-Aware Sparse Fine-Tuning Strategies for Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning large language models require expensive training on consumer-grade hardwares. |
| Approach: | They propose a sparsity-based approach that introduces trainable sparse adaptations to the weight matrices in the model and offers greater flexibility in selecting fine-tuned parameters. |
| Outcome: | The proposed method outperforms other methods for a simple yet effective baseline for nLP tasks while sacrificing performance. |
Efficient Many-Shot In-Context Learning with Dynamic Block-Sparse Attention (2025.acl-long)
Copied to clipboard
| Challenge: | Many-shot in-context learning shifts computational burden from training-time to inference-time, making deployment of many-shot ICL challenging to justify in-practice. |
| Approach: | They propose a method for retrieval-based many-shot in-context learning that uses blocks-sparse attention and retrieval of cached demonstrations to achieve comparable per-example latency to finetuning. |
| Outcome: | The proposed method achieves comparable per-example latency to finetuning while maintaining on average >95% of the best method’s accuracy across strong ICL and finetuned baselines. |
ScaleBiO: Scalable Bilevel Optimization for LLM Data Reweighting (2025.acl-long)
Copied to clipboard
Rui Pan, Dylan Zhang, Hanning Zhang, Xingyuan Pan, Minrui Xu, Jipeng Zhang, Renjie Pi, Xiaoyu Wang, Tong Zhang
| Challenge: | Existing paradigms for bilevel optimization require second-order information, making it difficult to scale them up. |
| Approach: | They propose a scalable instantiation of a bilevel optimization paradigm for large-scale LLMs by using a memory-efficient training technique. |
| Outcome: | The proposed paradigm scales to 30B-sized LLMs on 8H100 GPUs. |
PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference (2025.acl-long)
Copied to clipboard
Jiaming Ji, Donghai Hong, Borong Zhang, Boyuan Chen, Josef Dai, Boren Zheng, Tianyi Alex Qiu, Jiayi Zhou, Kaile Wang, Boxun Li, Sirui Han, Yike Guo, Yaodong Yang
| Challenge: | Using large-scale annotation data, large language models can generate noise, errors and biases, leading to unexpected behaviours. |
| Approach: | They propose a dataset to promote safety alignment in large language models . they separate helpfulness and harmlessness annotations for question-answering pairs . |
| Outcome: | The proposed dataset provides 44.6k prompts and 265k question-answer pairs with safety meta-labels for 19 harm categories and three severity levels, with answers generated by Llama-family models. |
What Happened in LLMs Layers when Trained for Fast vs. Slow Thinking: A Gradient Perspective (2025.acl-long)
Copied to clipboard
| Challenge: | Xu et al., 2024) study shows that slow thinking can distinguish correct and irrelevant reasoning paths. |
| Approach: | They investigate how fast vs. slow thinking affects layer-wise gradients in large language models . they find that slow thinking can distinguish correct and irrelevant reasoning paths . |
| Outcome: | The results show that slow thinking can distinguish correct and irrelevant reasoning paths. |
Beyond Text Compression: Evaluating Tokenizers Across Scales (2025.acl-long)
Copied to clipboard
| Challenge: | Language models rely on tokenizers to convert text into machine-interpretable tokens, which shape the statistical patterns that language models learn to estimate. |
| Approach: | They propose to use Zipf's law to measure tokenizer performance by combining several metrics to capture multiple aspects of tokenizer behavior. |
| Outcome: | The proposed metrics correlate more strongly with downstream performance than text compression when modeling unseen languages. |
Emergent Abilities of Large Language Models under Continued Pre-training for Language Adaptation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing large language models are notoriously English-centric, and their performance has been reported to drop significantly in lessresourced languages. |
| Approach: | They propose a language-agnostic benchmark for in-context learning that reveals catastrophic forgetting early on CPT when English is not included. |
| Outcome: | The proposed method does not impact validation perplexity but is critical for emergence of downstream capabilities in the target language. |
R-Fairness: Assessing Fairness of Ranking in Subjective Data (2025.acl-long)
Copied to clipboard
| Challenge: | Subjective data, reflecting individual opinions, permeates platforms like Yelp and Amazon . despite the prevalence of such platforms, little attention has been given to fairness in their context . |
| Approach: | They propose a fairness assessment pipeline that starts with data collection phase and then iterates through rated items. |
| Outcome: | The proposed approach favors groups writing best-ranked reviews over others on collaborative rating platforms. |
RePanda: Pandas-powered Tabular Verification and Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fact-checking tabular data rely on black-box models with opaque reasoning. |
| Approach: | They propose a structured fact verification approach that translates claims into executable pandas queries. |
| Outcome: | The proposed method outperforms existing methods and demonstrates strong OOD robustness. |
Towards Style Alignment in Cross-Cultural Translation (2025.acl-long)
Copied to clipboard
| Challenge: | Successful communication relies on the speaker’s intended style aligning with the listener’s interpreted style. |
| Approach: | They propose a method that leverages learned stylistic concepts to encourage LLM translation to appropriately convey cultural communication norms and align style. |
| Outcome: | The proposed method aims to encourage translations to convey cultural communication norms and align style. |
TiC-LM: A Web-Scale Benchmark for Time-Continual LLM Pretraining (2025.acl-long)
Copied to clipboard
Jeffrey Li, Mohammadreza Armandpour, Seyed Iman Mirzadeh, Sachin Mehta, Vaishaal Shankar, Raviteja Vemulapalli, Samy Bengio, Oncel Tuzel, Mehrdad Farajtabar, Hadi Pouransari, Fartash Faghri
| Challenge: | Large language models (LLMs) trained on historical web data inevitably become outdated. |
| Approach: | They propose a web-scale dataset for time-continual pretraining of LLMs derived from 114 dumps of Common Crawl (CC) they also design time-stratified evaluations to assess how well various continual learning methods adapt to new data while retaining past knowledge. |
| Outcome: | The proposed benchmarks show that autoregressive meta-schedules combined with a fixed-ratio replay of older data can achieve comparable held-out loss to re-training from scratch, while requiring significantly less computation (2.6x). |
Entailed Between the Lines: Incorporating Implication into NLI (2025.acl-long)
Copied to clipboard
Shreya Havaldar, Hamidreza Alvari, John Palowitch, Mohammad Javad Hosseini, Senaka Buthpitiya, Alex Fabrikant
| Challenge: | True Emotions, social cues, insults, and a myriad of other messages are conveyed implicitly, often even more so than explicitly. |
| Approach: | They propose a dataset to help LLMs understand implied entailment . |
| Outcome: | The proposed dataset enables LLMs to understand implied entailment and can generalize this understanding across datasets and domains. |
Multi-Level Explanations for Generative Language Models (2025.acl-long)
Copied to clipboard
Lucas Monteiro Paes, Dennis Wei, Hyo Jin Do, Hendrik Strobelt, Ronny Luss, Amit Dhurandhar, Manish Nagireddy, Karthikeyan Natesan Ramamurthy, Prasanna Sattigeri, Werner Geyer, Soumya Ghosh
| Challenge: | Large language models (LLMs) are being used for context-grounded tasks like summarizing meetings and answering doctors' questions. |
| Approach: | They propose a technique to provide explanations for context-grounded text generation by assigning scores to parts of the context to quantify their influence on the model output. |
| Outcome: | The proposed framework can provide more faithful explanations of generated output than available alternatives, including LLM self-explanations. |
A Multi-Agent Framework for Mitigating Dialect Biases in Privacy Policy Question-Answering Systems (2025.acl-long)
Copied to clipboard
Đorđe Klisura, Astrid R Bernaga Torres, Anna Karen Gárate-Escamilla, Rajesh Roshan Biswal, Ke Yang, Hilal Pataci, Anthony Rios
| Challenge: | Existing Privacy Policy Question Answering systems exhibit performance disparities across English dialects, disadvantaging speakers of non-standard varieties. |
| Approach: | They propose a framework that integrates a Dialect Agent and a Privacy Policy Agent to mitigate dialectal biases. |
| Outcome: | The proposed framework improves GPT-4o-mini’s zero-shot accuracy from 0.394 to 0.601 on PrivacyQA and 0.352 to 0.464 on PolicyQA. |
Low-Bit Quantization Favors Undertrained LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Larger models or those trained on fewer tokens exhibit less quantization-induced degradation (QiD), while smaller, well-trained models face significant performance losses. |
| Approach: | They propose to use QiD to measure an LLM’s training levels and determine the number of training tokens required for fully training LLMs of various sizes. |
| Outcome: | The proposed scaling laws can predict the quantization performance of different-sized LLMs trained with tokens. |
LETS-C: Leveraging Text Embedding for Time Series Classification (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in language modeling have shown promising results when applied to time series data. |
| Approach: | They propose a method to fine-tune large language models for time series classification tasks using text embedding models and a simple classification head. |
| Outcome: | The proposed model outperforms the current SOTA model on a time series classification benchmark and uses only 14.5% of the trainable parameters. |
UrbanVideo-Bench: Benchmarking Vision-Language Models on Embodied Intelligence with Video Data in Urban Spaces (2025.acl-long)
Copied to clipboard
Baining Zhao, Jianjie Fang, Zichao Dai, Ziyou Wang, Jirong Zha, Weichen Zhang, Chen Gao, Yue Wang, Jinqiang Cui, Xinlei Chen, Yong Li
| Challenge: | Large multimodal models exhibit remarkable intelligence, yet their embodied cognitive abilities during motion in open-ended urban aerial spaces remain to be explored. |
| Approach: | They propose a benchmark to evaluate whether large multimodal models can process continuous first-person visual observations like humans. |
| Outcome: | The proposed model can process first-person visual observations like humans, enabling recall, perception, reasoning, and navigation. |
HELIOS: Harmonizing Early Fusion, Late Fusion, and LLM Reasoning for Multi-Granular Table-Text Retrieval (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for table-text retrieval are limited due to the need to bridge structured tables and unstructured passages. |
| Approach: | They propose a table-text retrieval system that combines the strengths of both approaches . they propose bipartite subgraph retrieval and query-relevant node expansion . |
| Outcome: | The proposed method outperforms state-of-the-art models with a 42.6% and 39.9% improvement on the OTT-QA benchmark. |
ONEBench to Test Them All: Sample-Level Benchmarking Over Open-Ended Capabilities (2025.acl-long)
Copied to clipboard
| Challenge: | ONEBench enables custom benchmarks for specific capabilities while reusing and aggregating samples. |
| Approach: | They propose a new paradigm that consolidates individual evaluation datasets into a unified, ever-expanding sample pool. |
| Outcome: | The proposed model evaluation framework is based on dynamic, sample-level evaluation. |
La Leaderboard: A Large Language Model Leaderboard for Spanish Varieties and Languages of Spain and Latin America (2025.acl-long)
Copied to clipboard
María Grandury, Javier Aula-Blasco, Júlia Falcão, Clémentine Fourrier, Miguel González Saiz, Gonzalo Martínez, Gonzalo Santamaria Gomez, Rodrigo Agerri, Nuria Aldama García, Luis Chiruzzo, Javier Conde, Helena Gomez Adorno, Marta Guerrero Nieto, Guido Ivetta, Natàlia López Fuertes, Flor Miriam Plaza-del-Arco, María-Teresa Martín-Valdivia, Helena Montoro Zamorano, Carmen Muñoz Sanz, Pedro Reviriego, Leire Rosado Plaza, Alejandro Vaca Serrano, Estrella Vallecillo-Rodríguez, Jorge Vallego, Irune Zubiaga
| Challenge: | La Leaderboard is the first open-source leaderboard to evaluate generative Large Language Models (LLMs) in languages and language varieties of Spain and Latin America. |
| Approach: | They propose to use La Leaderboard to evaluate generative Large Language Models in Spanish and Latin America. |
| Outcome: | La Leaderboard is the first open-source leaderboard to evaluate generative LLMs in languages and language varieties of Spain and Latin America. |
Why Prompt Design Matters and Works: A Complexity Analysis of Prompt Search Space in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Chain-of-Thought (CoT) prompting has emerged as a practical workaround, but most CoT-based methods rely on a single generic prompt like “think step by step” with no task-specific adaptation. |
| Approach: | They propose a theoretical framework that explains why some prompts succeed while others fail by using a generic generic prompt like "think step by step" they show that prompts function as selectors, extracting specific task-relevant information from the model's full hidden state during CoT reasoning. |
| Outcome: | The proposed framework explains why some prompts succeed while others fail. |
Energy Considerations of Large Language Model Inference and Efficiency Optimizations (2025.acl-long)
Copied to clipboard
| Challenge: | Prior benchmarking efforts focused on latency reduction in idealized settings, often overlooking real-world inference workloads that shape energy use. |
| Approach: | They propose a modeling approach that approximates real-world LLM workflows . they show that the effectiveness of inference optimizations is sensitive to workload geometry . |
| Outcome: | The proposed approach reduces energy use by 73% from unoptimized baselines. |
Optimizing Pre-Training Data Mixtures with Mixtures of Data Expert Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to optimize language model pre-training data mixtures are difficult due to the complexity of the data mixture. |
| Approach: | They propose a method to optimize language model pre-training data mixtures by approximating cross-entropy loss via a Mixture of Data Experts (MDE). |
| Outcome: | The proposed method improves performance on a slimPajama dataset with a mixture of data experts. |
BFS-Prover: Scalable Best-First Tree Search for LLM-based Automatic Theorem Proving (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to theorem proving in large language models rely on value functions and/or Monte Carlo Tree Search (MCTS), but the potential of simpler methods like Best-First Tree Search remains underexplored. |
| Approach: | They propose a scalable expert iteration framework that implements strategic data filtering at each expert iteration round, excluding problems solvable via beam search node expansion to focus on harder cases. |
| Outcome: | The proposed framework achieves a state-of-the-art score of 72.95 on the MiniF2F test set and challenges the perceived necessity of complex tree search methods. |
Magnet: Multi-turn Tool-use Data Synthesis and Distillation via Graph Translation (2025.acl-long)
Copied to clipboard
Fan Yin, Zifeng Wang, I-Hung Hsu, Jun Yan, Ke Jiang, Yanfei Chen, Jindong Gu, Long Le, Kai-Wei Chang, Chen-Yu Lee, Hamid Palangi, Tomas Pfister
| Challenge: | Large language models have been shown to be effective in multi-turn interactions . however, their performance may be limited in complex, multi-turned interactions involving users and multiple tools. |
| Approach: | They propose a framework for synthesizing high-quality training trajectories to enhance the function calling capability of large language model agents in multi-turn conversations with humans. |
| Outcome: | The proposed model outperforms the teacher model by 68.01 on BFCL-v3 and 73.30 on ToolQuery. |
Logic-Regularized Verifier Elicits Reasoning from LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Typical verifiers require resource-intensive supervised dataset construction, which is costly and faces limitations in data diversity. |
| Approach: | They propose an unsupervised verifier regularized by logical rules that uses internal activations and logical constraints on multiple reasoning paths. |
| Outcome: | Experiments on 10 datasets show that the proposed verifier outperforms baselines and is comparable to the supervised verifier. |
Squeezed Attention: Accelerating Long Context Length LLM Inference (2025.acl-long)
Copied to clipboard
Coleman Richard Charles Hooper, Sehoon Kim, Hiva Mohammadzadeh, Monishwaran Maheswaran, Sebastian Zhao, June Paik, Michael W. Mahoney, Kurt Keutzer, Amir Gholami
| Challenge: | Emerging Large Language Models require long input context to perform complex tasks. |
| Approach: | They propose an algorithm to reduce the complexity of attention with respect to the fixed context length. |
| Outcome: | The proposed method reduces the complexity of attention from linear to logarithmic with respect to the fixed context length. |
LangMark: A Multilingual Dataset for Automatic Post-Editing (2025.acl-long)
Copied to clipboard
Diego Velazquez, Mikaela Grace, Konstantinos Karageorgos, Lawrence Carin, Aaron Schliem, Dimitrios Zaikis, Roger Wechsler
| Challenge: | Automated post-editing (APE) aims to correct errors in machine-translated text . lack of large-scale multilingual datasets specifically tailored to NMT outputs hinders APE development . |
| Approach: | They propose to use a human-annotated multilingual APE dataset for English translation to seven languages to address this gap. |
| Outcome: | The proposed dataset offers both linguistic diversity and scale. |
Neural Parameter Search for Slimmer Fine-Tuned Models and Better Transfer (2025.acl-long)
Copied to clipboard
Guodong Du, Zitao Fang, Jing Li, Junlin Li, Runhua Jiang, Shuyang Yu, Yifei Guo, Yangneng Chen, Sim Kuan Goh, Ho-Kin Tang, Daojing He, Honghai Liu, Min Zhang
| Challenge: | Foundational models and their checkpoints have advanced deep learning, boosting performance across applications. |
| Approach: | They propose a method for pruning fine-tuned models by calculating differences between them and original model. |
| Outcome: | The proposed method can improve performance across vision, NLP, and multi-modal benchmarks. |
Merge Hijacking: Backdoor Attacks to Model Merging of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing research on model merging focuses on optimizing model performance and minimizing backdoors. |
| Approach: | They propose a backdoor attack targeting model merging in Large Language Models that creates a unified model for multi-domain tasks. |
| Outcome: | The proposed attack is effective across models, merging algorithms, and tasks while maintaining utility across tasks. |
Where Are We? Evaluating LLM Performance on African Languages (2025.acl-long)
Copied to clipboard
Ife Adebara, Hawau Olamide Toyin, Nahom Tesfu Ghebremichael, AbdelRahim A. Elmadany, Muhammad Abdul-Mageed
| Challenge: | African languages are underrepresented in NLP due to policies that favor foreign languages and create data inequities. |
| Approach: | They integrate theoretical insights on Africa’s language landscape with an empirical evaluation using Sahara datasets. |
| Outcome: | The proposed model improves on a benchmark curated from large-scale, publicly accessible datasets capturing the continent's linguistic diversity. |
Beyond Output Matching: Bidirectional Alignment for Enhanced In-Context Learning (2025.acl-long)
Copied to clipboard
Chengwei Qin, Wenhan Xia, Fangkai Jiao, Chen Chen, Yuchen Hu, Bosheng Ding, Ruirui Chen, Shafiq Joty
| Challenge: | Existing methods to train student models on the generated outputs of teacher models are not efficient for ICL. |
| Approach: | They propose to align the output of smaller (student) models with that of larger (teacher) models by incorporating a ranking loss and aligning the token-level output distribution. |
| Outcome: | The proposed model outperforms baseline models on a variety of tasks involving language understanding, reasoning, and coding. |
CiteEval: Principle-Driven Citation Evaluation for Source Attribution (2025.acl-long)
Copied to clipboard
Yumo Xu, Peng Qi, Jifan Chen, Kunlun Liu, Rujun Han, Lan Liu, Bonan Min, Vittorio Castelli, Arshit Gupta, Zhiguo Wang
| Challenge: | Current evaluation frameworks rely on NLI to assess binary or ternary support from cited sources, which is suboptimal for citation evaluation. |
| Approach: | They propose a citation evaluation framework based on fine-grained citation ratings within a broad context and construct a multi-domain benchmark with high-quality human annotations. |
| Outcome: | The proposed framework provides a high-quality human annotation benchmark and a suite of model-based metrics that exhibit strong correlation with human judgments. |
HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to optimize agent performance by incorporating entire historical action-observation pairs into LLMs are redundant in long-horizon tasks. |
| Approach: | They propose a framework that leverages subgoals as memory chunks to manage working memory of LLM-based agents hierarchically. |
| Outcome: | The proposed framework achieves a twofold increase in success rate and reduces the average number of steps required by 3.8. |
EducationQ: Evaluating LLMs’ Teaching Capabilities Through Multi-Agent Dialogue Framework (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly used as educational tools, yet evaluating their teaching capabilities remains challenging due to the resource-intensive nature of teacher-student interactions. |
| Approach: | They propose a multi-agent dialogue framework that efficiently assesses teaching capabilities through simulated dynamic educational scenarios. |
| Outcome: | The proposed framework outperforms open-source models on 1,498 questions across 13 disciplines and 10 difficulty levels on 1,400 questions. |
KRISTEVA: Close Reading as a Novel Task for Benchmarking Interpretive Reasoning (2025.acl-long)
Copied to clipboard
Peiqi Sui, Juan Diego Rodriguez, Philippe Laban, J. Dean Murphy, Joseph P. Dexter, Richard Jean So, Samuel Baker, Pramit Chaudhuri
| Challenge: | a study of close reading skills in large language models (LLMs) shows that LLMs still lag behind human evaluators on 10 of 11 tasks. |
| Approach: | They propose a benchmark to evaluate close reading skills in large language models . they propose three tasks to approximate different elements of the close reading process . |
| Outcome: | The proposed benchmarks show that state-of-the-art LLMs possess some college-level close reading competency, but performance still trails human evaluators on 10 out of 11 tasks. |
Efficient Domain Continual pretraining by Mitigating the Stability Gap (2025.acl-long)
Copied to clipboard
| Challenge: | Continual pretraining is an important approach for Large Language Models to improve their performance in target domains, learn new topics and languages, and even boost their general capabilities. |
| Approach: | They propose a training strategy that mitigates instability by increasing the number of epochs, along with two data sampling strategies targeting data domain relevance and corpus distribution. |
| Outcome: | The proposed training strategy improves the average medical task performance of the OpenLlama-3B model from 36.2% to 40.7% using only 40% of the original training budget, while also enhancing general task performance without causing forgetting. |
Palm: A Culturally Inclusive and Linguistically Diverse Dataset for Arabic LLMs (2025.acl-long)
Copied to clipboard
Fakhraddin Alwajih, Abdellah El Mekki, Samar Mohamed Magdy, AbdelRahim A. Elmadany, Omer Nacar, El Moatez Billah Nagoudi, Reem Abdel-Salam, Hanin Atwany, Youssef Nafea, Abdulfattah Mohammed Yahya, Rahaf Alhamouri, Hamzah A. Alsayadi, Hiba Zayed, Sara Shatnawi, Serry Sibaee, Yasir Ech-chammakhy, Walid Al-Dhabyani, Marwa Mohamed Ali, Imen Jarraya, Ahmed Oumar El-Shangiti, Aisha Alraeesi, Mohammed Anwar AL-Ghrawi, Abdulrahman S. Al-Batati, Elgizouli Mohamed, Noha Taha Elgindi, Muhammed Saeed, Houdaifa Atou, Issam Ait Yahia, Abdelhak Bouayad, Mohammed Machrouh, Amal Makouar, Dania Alkawi, Mukhtar Mohamed, Safaa Taher Abdelfadil, Amine Ziad Ounnoughene, Anfel Rouabhia, Rwaa Assi, Ahmed Sorkatti, Mohamedou Cheikh Tourad, Anis Koubaa, Ismail Berrada, Mustafa Jarrar, Shady Shehata, Muhammad Abdul-Mageed
| Challenge: | a year-long community-driven project covering all 22 Arab countries evaluates the cultural and dialectal capabilities of large language models. |
| Approach: | They propose a project to evaluate the cultural and dialectal capabilities of large language models. |
| Outcome: | The project evaluates the cultural and dialectal capabilities of several frontier LLMs. |
NewsInterview: a Dataset and a Playground to Evaluate LLMs’ Grounding Gap via Informational Interviews (2025.acl-long)
Copied to clipboard
Alexander Spangher, Michael Lu, Sriya Kalyan, Hyundong Justin Cho, Tenghao Huang, Weiyan Shi, Jonathan May
| Challenge: | Existing large datasets (1k-10k transcripts) are generated via crowdsourcing and are inherently unnatural. |
| Approach: | They curate a dataset of 40,000 two-person informational interviews from NPR and CNN . they find that LLMs are significantly less likely than human interviewers to use acknowledgements and pivot to higher-level questions. |
| Outcome: | The proposed model is based on 40,000 interviews with journalists and CNN . |
CFBench: A Comprehensive Constraints-Following Benchmark for LLMs (2025.acl-long)
Copied to clipboard
Tao Zhang, ChengLIn Zhu, Yanjun Shen, Wenjing Luo, Yan Zhang, Hao Liang, Tao Zhang, Fan Yang, Mingan Lin, Yujing Qiao, Weipeng Chen, Bin Cui, Wentao Zhang, Zenan Zhou
| Challenge: | Existing evaluations of Large Language Models (LLMs) focus on fragmented constraints or narrow scenarios, but they overlook the comprehensiveness and authenticity of constraints from the user’s perspective. |
| Approach: | They propose a Chinese Comprehensive Constraints Following Benchmark for LLMs that compiles constraints from real-world instructions and constructs a systematic framework for constraint types. |
| Outcome: | The proposed framework integrates multi-dimensional assessment criteria with requirement prioritization, covering various perspectives of constraints, instructions, and requirement fulfillment. |
Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages (2025.acl-long)
Copied to clipboard
Ashwin Sankar, Sparsh Jain, Nikhil Narasimhan, Devilal Choudhary, Dhairya Suman, Mohammed Safi Ur Rahman Khan, Anoop Kunchukuttan, Mitesh M Khapra, Raj Dabre
| Challenge: | Existing datasets that cover only a fraction of Indian languages lack the breadth needed to generalize beyond curated benchmarks. |
| Approach: | They propose to build the largest speech translation dataset for Indian languages . they use a three-step methodology to gather data and train a model that performs better . |
| Outcome: | The proposed model improves on existing models and is open-source with permissive licenses. |
CoRe-MMRAG: Cross-Source Knowledge Reconciliation for Multimodal RAG (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Multimodal Large Language Models have significantly improved reasoning and generation tasks by leveraging joint vision-language representations. |
| Approach: | They propose a framework that reconciles inconsistencies across knowledge sources . they use a four-stage pipeline to generate an internal response from parametric knowledge . |
| Outcome: | Experiments on KB-VQA show that CoRe-MMRAG achieves performance gains of 5.6% and 9.3% over baseline methods. |
Mapping 1,000+ Language Models via the Log-Likelihood Vector (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods to compare autoregressive language models are based on log-likelihoods . a model map is constructed using coordinates that capture the geometric structure of probability distributions based upon text-generation probabilities. |
| Approach: | They propose to use log-likelihood vectors to compare autoregressive language models . when treated as model features, their squared Euclidean distance approximates KL divergence . |
| Outcome: | The proposed method is highly scalable and easy to implement. |
ConsistencyChecker: Tree-based Evaluation of LLM Generalization Capabilities (2025.acl-long)
Copied to clipboard
| Challenge: | Traditional self-consistency methods fail to capture subtle semantic errors in multi-step tasks. |
| Approach: | They propose a tree-based evaluation framework that measures LLMs’ ability to preserve semantic consistency during reversible transformations. |
| Outcome: | The proposed framework measures generalization abilities across models from 1.5B to 72B and can be used to benchmark LLMs without constructing new datasets. |
Robust Estimation of Population-Level Effects in Repeated-Measures NLP Experimental Designs (2025.acl-long)
Copied to clipboard
| Challenge: | a study shows that linear mixed-effects models can capture language-dependent differences in a population . a long-standing concern in experimental linguistics is the multilevel nature of language data . significance testing is inconsistent or improperly applied, authors say . |
| Approach: | a case study shows that linear mixed-effects models can capture language-dependent differences . authors encourage NLP community to embrace variability as feature, not nuisance . a long-standing concern in experimental linguistics is multilevel, hierarchical nature of language data . |
| Outcome: | a case study shows that mixed-effects models can capture language-dependent differences . authors encourage the NLP community to embrace variability as a feature, not a nuisance . |
FactBench: A Dynamic Benchmark for In-the-Wild Language Model Factuality Evaluation (2025.acl-long)
Copied to clipboard
| Challenge: | Language models (LMs) generate false or unverifiable content, often known as hallucination, despite ongoing efforts to enhance their factuality. |
| Approach: | They propose a tool that measures LMs’ factuality in real-world user interactions by evaluating their factual accuracy and categorizing content units as Supported, Unsupported, or Undecidable based on Web-retrieved evidence. |
| Outcome: | The proposed evaluation pipeline measures language models’ factuality in real-world user interactions. |
Training-free LLM Merging for Multi-task Learning (2025.acl-long)
Copied to clipboard
Zichuan Fu, Xian Wu, Yejing Wang, Wanyu Wang, Shanshan Ye, Hongzhi Yin, Yi Chang, Yefeng Zheng, Xiangyu Zhao
| Challenge: | Large Language Models (LLMs) have demonstrated exceptional capabilities across diverse natural language processing tasks. |
| Approach: | They propose a training-free method for unifying different specialized LLMs into a single model using model-wise and layer-wise pruning and scaling. |
| Outcome: | The proposed method outperforms existing merging techniques and surpasses models fine-tuned on combined datasets in most scenarios. |
Inferring from Logits: Exploring Best Practices for Decoding-Free Generative Candidate Selection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing work has been using decoding-free candidate selection methods to obtain candidate probability from initial output logits over vocabulary. |
| Approach: | They propose to evaluate a set of tasks using decoding-free candidate selection methods on a comprehensive set of questions. |
| Outcome: | The proposed methods are evaluated on a set of tasks including five multiple-choice QA tasks with a small candidate pool and four clinical decision tasks with 10k+ options. |
Comparison-based Active Preference Learning for Multi-dimensional Personalization (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models have shown remarkable success, but aligning them with human preferences remains a core challenge. |
| Approach: | They propose to capture implicit user preferences from comparative feedback to improve model performance. |
| Outcome: | The proposed framework is able to capture implicit user preferences from comparative feedback. |
OpenCoder: The Open Cookbook for Top-Tier Code Large Language Models (2025.acl-long)
Copied to clipboard
Siming Huang, Tianhao Cheng, Jason Klein Liu, Weidi Xu, Jiaran Hao, Liuyihan Song, Yang Xu, Jian Yang, Jiaheng Liu, Chenchen Zhang, Linzheng Chai, Ruifeng Yuan, Xianzhen Luo, Qiufeng Wang, YuanTao Fan, Qingfu Zhu, Zhaoxiang Zhang, Yang Gao, Jie Fu, Qian Liu, Houyi Li, Ge Zhang, Yuan Qi, Xu Yinghui, Wei Chu, Zili Wang
| Challenge: | Code LLMs lack reproducible data pipelines and training protocols for reproducible advancements in code intelligence. |
| Approach: | They propose a top-tier code LLM that releases model weights and inference code . reproducible data pipelines, rigorous experimental ablation results and training protocols are included . |
| Outcome: | The proposed model achieves comparable performance to leading models and serves as an "open cookbook" reproducible training data, rigorous experimental ablation results, and detailed training protocols are also included in the model. |
LlamaDuo: LLMOps Pipeline for Seamless Migration from Service LLMs to Small-Scale Local LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | LLMOps pipelines are used to migrate knowledge and abilities from service-oriented LLMs to smaller, locally manageable models. |
| Approach: | They propose an LLMOps pipeline for the seamless migration of knowledge and abilities from service-oriented LLMs to smaller, locally manageable models. |
| Outcome: | Experiments with leading-edge LLMs show that the proposed pipeline can scale to meet various tasks and domains. |
AmbiK: Dataset of Ambiguous Tasks in Kitchen Environment (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are used for behavior planning given natural language instructions from the user. |
| Approach: | They propose to use a textual dataset of ambiguous instructions addressed to a robot in a kitchen environment to compare them. |
| Outcome: | The proposed dataset includes 1000 pairs of ambiguous tasks and their unambiguous counterparts, with environment descriptions, clarifying questions and answers, user intents, and task plans. |
SocialCC: Interactive Evaluation for Cultural Competence in Language Agents (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies have evaluated cultural knowledge of large language models, but they fail to assess dynamic cultural competence. |
| Approach: | They propose a benchmark to assess cultural competence through intercultural scenarios that span 60 countries across six continents. |
| Outcome: | The proposed benchmark measures the ability of large language models to apply cultural knowledge effectively in cross-cultural interactions. |
Scalable Vision Language Model Training via High Quality Data Curation (2025.acl-long)
Copied to clipboard
| Challenge: | SAIL-VL models achieve the highest average score in 18 widely used VLM benchmarks in our evaluation, with the 2B model takes the top position over VLMs of comparable sizes on OpenCompass 2024. |
| Approach: | They introduce an open-source vision language model (VLM) series that can be trained using high-quality data. |
| Outcome: | The proposed model achieves the highest average score in 18 widely used VLM benchmarks, with the 2B model taking the top position over VLMs of comparable sizes on OpenCompass 2024. |
GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusion (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies rely on item metadata to construct abbreviated item IDs, leading to a loss of valuable details. |
| Approach: | They propose a Generative Recommender via semantic-aware multi-granular late fusion to integrate rich semantics efficiently with minimal information loss. |
| Outcome: | The proposed model outperforms eight state-of-the-art recommendation models on four benchmark datasets and achieves significant improvements of 11.5-16.0% in Recall@5 and 5.3-13.6% in NDCG@5. |
Towards Economical Inference: Enabling DeepSeek’s Multi-Head Latent Attention in Any Transformer-based LLMs (2025.acl-long)
Copied to clipboard
Tao Ji, Bin Guo, Yuanbin Wu, Qipeng Guo, Shenlixing Shenlixing, Chenzhan Chenzhan, Xipeng Qiu, Qi Zhang, Tao Gui
| Challenge: | Multi-head Latent Attention (MLA) is an innovative architecture designed to ensure efficient and economical inference by significantly compressing the Key-Value (KV) cache into a latent vector. |
| Approach: | They propose a data-efficient fine-tuning method for transitioning from MHA to MLA using a latent vector cache. |
| Outcome: | The proposed architecture reduces the KV cache size of Llama2-7B by 92.19%, with only 1% drop in LongBench performance. |
TETRIS: Optimal Draft Token Selection for Batch Speculative Decoding (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods that optimize for a single request or a group of requests as a whole only select the most promising draft tokens to be accepted when verified in parallel. |
| Approach: | They propose a method that optimizes the total throughput of batch speculative decoding in multi-request settings by actively selecting the most promising draft tokens to be accepted when verified in parallel. |
| Outcome: | The proposed method outperforms baseline speculative decoding and existing methods that dynamically select draft tokens, leading to a more efficient batch inference in large language models. |
Introducing Verification Task of Set Consistency with Set-Consistency Energy Networks (2025.acl-long)
Copied to clipboard
| Challenge: | Traditional methods that rely on 1:1 pairwise comparisons fail to capture inconsistencies . few studies have addressed related problems in the domain of factual inconsistenency detection . |
| Approach: | They propose a set-consistency verification task that assesses logical coherence of entire sets . they propose 'set-consistent energy network' that employs a margin-based loss to learn the compatibility among a collection of statements . |
| Outcome: | The proposed model outperforms existing methods and significantly outperformed existing models. |
Language Models can Subtly Deceive Without Lying: A Case Study on Strategic Phrasing in Legislation (2025.acl-long)
Copied to clipboard
Atharvan Dogra, Krishna Pillutla, Ameet Deshpande, Ananya B. Sai, John J Nay, Tanmay Rajpurohit, Ashwin Kalyan, Balaraman Ravindran
| Challenge: | blatant lying or unintentional hallucination are common in large language models. |
| Approach: | They build a testbed mimicking a legislative environment where a corporate lobbyist module is proposing amendments to bills that benefit a specific company while evading identification by strong LLM detectors. |
| Outcome: | The proposed model can be used to detect deception in legislative environments and to optimize its phrasing to avoid detection by strong detectors. |
AfroCS-xs: Creating a Compact, High-Quality, Human-Validated Code-Switched Dataset for African Languages (2025.acl-long)
Copied to clipboard
Kayode Olaleye, Arturo Oncevay, Mathieu Sibue, Nombuyiselo Zondi, Michelle Terblanche, Sibongile Mapikitla, Richard Lastrucci, Charese Smiley, Vukosi Marivate
| Challenge: | AfroCS-xs is a low-quality dataset for code-switching in multilingual communities . code-witching is prevalent in multicultural societies but lacks high-quality data for model development . |
| Approach: | They propose to use human-validated synthetic code-switched datasets to generate code-witched sentences for four African languages and English within a specific domain—agriculture. |
| Outcome: | The proposed model improves translation accuracy on the high-quality dataset for four African languages and English within a specific domain—agriculture. |
Just Go Parallel: Improving the Multilingual Capabilities of Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have impressive translation capabilities even without being explicitly trained on parallel data. |
| Approach: | They propose to add parallel data to enhance multilingual encoder-based and encoder decoder language models by focusing on translation and multilingual common-sense reasoning. |
| Outcome: | The proposed methods show that adding parallel data can significantly improve LLMs’ multilingual capabilities. |
Design Choices for Extending the Context Length of Visual Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing open-source Visual Language Models lack systematic exploration into extending their context length, and commercial models often provide limited details. |
| Approach: | They propose to extend Visual Language Models (VLMs) to 128K lengths and open-source the code, data, and models. |
| Outcome: | The proposed model is based on the Qwen-VL series model and is competitive with commercial model GPT-4V. |