Papers by Wang Cai
Non-Autoregressive Text Generation with Pre-trained Language Models (2021.eacl-main)
Copied to clipboard
| Challenge: | Autoregressive generation models generate tokens in a left-to-right, token-by-token fashion, resulting in lag in inference. |
| Approach: | They propose to use BERT as the backbone of a non-autoregressive generation model for greatly improved performance. |
| Outcome: | The proposed model outperforms existing non-autoregressive models and achieves competitive performance with many strong autoregressive model. |
LongLLaVA: Scaling Multi-modal LLMs to 1000 Images Efficiently via a Hybrid Architecture (2025.findings-emnlp)
Copied to clipboard
Xidong Wang, Dingjie Song, Shunian Chen, Junying Chen, Zhenyang Cai, Chen Zhang, Lichao Sun, Benyou Wang
| Challenge: | Long-context Large Language Models (MLLMs) are critical for video understanding and image analysis. |
| Approach: | They propose a hybrid architecture that integrates Mamba and Transformer blocks . they introduce data construction methods that capture both temporal and spatial dependencies . |
| Outcome: | The proposed model achieves competitive results across various benchmarks while maintaining high throughput and low memory consumption. |
ASPO: Adaptive Sentence-Level Preference Optimization for Fine-Grained Multimodal Reasoning (2025.findings-acl)
Copied to clipboard
| Challenge: | Recent advances have extended DPO to multimodal scenarios, achieving strong performance. |
| Approach: | They propose to use a sentence-level preference optimization technique to optimize individual sentences for more precise preference optimization without additional models or parameters. |
| Outcome: | Experiments show that Adaptive Sentence-level Preference Optimization significantly improves the alignment of multimodal models. |
OSC: Cognitive Orchestration through Dynamic Knowledge Alignment in Multi-Agent LLM Collaboration (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Prior work has advanced agent selection and result aggregation, efficient linguistic interactions for deep collaboration among expert agents remain a critical bottleneck. |
| Approach: | They propose a knowledge-aware adaptive collaboration framework to enhance cognitive synergy in multi-agent systems with large language models. |
| Outcome: | The proposed framework improves synergy between agents and language models by enabling agents to dynamically perceive their collaborators’ cognitive states. |
K-order Ranking Preference Optimization for Large Language Models (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing list-wise methods focus on optimizing list ranking consistency for LLMs to improve ranking abilities. |
| Approach: | They propose to extend the Plackett-Luce model to accommodate top-K ranking by extending the DPO’s Plact-Lucer model to dynamically determine appropriate K for different samples. |
| Outcome: | The proposed model can be extended to accommodate top-K ranking and improve training efficiency. |
Make Your Decision Convincing! A Unified Two-Stage Framework: Self-Attribution and Decision-Making (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing frameworks for explaining black-box model behavior are unreliable . large-scale pre-trained models often rely on superficial clues for predictions . |
| Approach: | They propose a unified two-stage framework that uses subsequences from the input text as a rationale to generate model decision. |
| Outcome: | The proposed framework achieves competitive results on five reasoning datasets and in semi-supervised scenarios. |
Thinking with DistilQwen: A Tale of Four Distilled Reasoning and Reward Model Series (2025.emnlp-industry)
Copied to clipboard
| Challenge: | In the rapidly evolving landscape of large language models, the need for efficient reasoning models has become increasingly urgent. |
| Approach: | They extend the Qwen model family by introducing four model series specifically designed for industrial applications. |
| Outcome: | The proposed models outperform previous models in multiple benchmarks and provide scalable training and inference functionality on the Alibaba Cloud PAI platform. |
GraphCache: Message Passing as Caching for Sentence-Level Relation Extraction (2022.findings-naacl)
Copied to clipboard
| Challenge: | Existing work only encodes entity types and textual context within individual instances, which limits the performance of sentence-level relation extraction (RE). |
| Approach: | They propose a module that aggregates the features from sentences to learn global representations of properties and augments local features within individual sentences. |
| Outcome: | The proposed module can learn global representations of properties from sentences and augment local features within individual sentences. |
A Logical Pattern Memory Pre-trained Model for Entailment Tree Generation (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing models overlook the importance of generating intermediate conclusions with logical consistency from the given facts, leading to inaccurate conclusions and undermining the overall credibility of entailment trees. |
| Approach: | They propose a model that utilizes logical entailment patterns to generate coherent explanations by leveraging logical patterns. |
| Outcome: | The proposed model produces more coherent and reasonable conclusions that closely align with the underlying premises. |
Updating Large Language Models’ Memories with Time Constraints (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models (LLMs) can modify their internal memory by incorporating the latest external knowledge, but in practical applications, outdated information may be inputted into LLMs. |
| Approach: | They propose a two-stage decoupling framework that separates the identification and computation of time constraints into a symbolic system and propose 'selective update' of internal memory based on time constraints. |
| Outcome: | The proposed framework improves ChatGPT performance by 60% and improves state-of-the-art LLM GPT-4. |
Reinforcement Learning for Diffusion LLMs via Energy-Based Gibbs Alignment (2026.acl-long)
Copied to clipboard
| Challenge: | Diffusion Large Language Models (dLLMs) offer parallel decoding and bidirectional context modeling . aligning dLLms with reinforcement learning (RL) remains a challenge . |
| Approach: | They propose a variational framework that reformulates RL for dLLMs as a distribution matching problem. |
| Outcome: | The proposed framework reformulates RL for dLLMs as a distribution matching problem. |
DentalGPT: Incentivizing Multimodal Reasoning in Dentistry (2026.findings-acl)
Copied to clipboard
Zhenyang Cai, Jiaming Zhang, Junjie Zhao, Ziyi Zeng, Yanchao Li, Liang Jingyi, Junying Chen, Yunjin Yang, Jiajun You, Shuzhi Deng, null Xieruiqiii, Yuanting Chen, Xiangyi Feng, Jianquan Li, Liangyi Chen, Junwen Wang, Shan Jiang, Benyou Wang
| Challenge: | Current multimodal large language models (MLLMs) show limited understanding of dental images. |
| Approach: | They propose a dental-specialized multimodal large language model trained via staged multimodal alignment and reinforcement learning. |
| Outcome: | The proposed model outperforms state-of-the-art models on disease classification and dental VQA tasks. |
Mapping the Minds of LLMs: A Graph-Based Analysis of Reasoning LLMs (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large Reasoning Models (LRMs) often display unstable behaviors, e.g., hallucinating unsupported premises, overthinking simple tasks, and displaying higher sensitivity to prompt variations. |
| Approach: | They propose a graph-based analytical framework that clusters long, verbose CoT outputs into semantically coherent reasoning steps, then constructs directed reasoning graphs to capture contextual and logical dependencies among these steps. |
| Outcome: | The proposed framework enables quantitative evaluation of internal reasoning structure and quality beyond conventional metrics and provides practical insights for prompt engineering and cognitive analysis of LLMs. |
Enhancing Language Representation with Constructional Information for Natural Language Understanding (2023.acl-long)
Copied to clipboard
| Challenge: | Recent advances in natural language processing focus on acquiring lexico-semantic information. |
| Approach: | They propose a construction grammar which highlights the pairings of form and meaning to enrich language representation. |
| Outcome: | The proposed model is superior to existing models on a variety of NLU tasks. |
Generative-to-Discriminative Test-Time Adaptation via Manifold-Aware Diffusion and Bayesian Distillation (2026.findings-acl)
Copied to clipboard
Boyun Zhang, Zequn Xie, Fangming Feng, Zihan Zhang, Yongbo He, Chuxin Wang, Sihang Cai, Tao Jin, Qifei Zhang
| Challenge: | Existing discriminative approaches suffer from "confident but wrong" failure mode, blindly adapting to OOD noise leading to error accumulation. |
| Approach: | They propose a TTA framework that harmonizes the robustness of generative diffusion models with the efficiency of discriminative regression networks via Bayesian Diffusion Distillation (BDD). |
| Outcome: | The proposed framework reduces MAE from 0.6872 to 0.5673 and boosts binary accuracy by 5.81 percentage points (reaching 57.33%) it also reduces the MAE of the MOSI to SIMS shift and achieves an 11.18-point gain over the baseline. |
UnifiedMLLM: Enabling Unified Representation for Multi-modal Multi-tasks With Large Language Model (2025.findings-naacl)
Copied to clipboard
Zhaowei Li, Wei Wang, YiQing Cai, Qi Xu, Pengyu Wang, Dong Zhang, Hang Song, Botian Jiang, Zhida Huang, Tao Wang
| Challenge: | Representative models like LLaVA and MiniGPT-4 have great capabilities in various tasks. |
| Approach: | They propose a unified model to represent various multi-modal tasks using a single representation. |
| Outcome: | The proposed model outperforms existing models in a variety of tasks while maintaining generality and scalability. |
Incorporating Circumstances into Narrative Event Prediction (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Existing studies focus on mining the inter-events relationships while ignoring how the events happened. |
| Approach: | They propose to incorporate event circumstances into the narrative event prediction by combining two multi-head attention modules and regularizing attention weights. |
| Outcome: | The proposed model outperforms baseline models by 12.2%. |
Large Language Models are not Fair Evaluators (2024.acl-long)
Copied to clipboard
Peiyi Wang, Lei Li, Liang Chen, Zefan Cai, Dawei Zhu, Binghuai Lin, Yunbo Cao, Lingpeng Kong, Qi Liu, Tianyu Liu, Zhifang Sui
| Challenge: | Existing evaluation frameworks that use large language models as referees are insufficient for accurately assessing their alignment with human intent. |
| Approach: | They propose a calibration framework to address positional bias in large language models as evaluators by manually annotating the “win/tie/lose” outcomes of responses from ChatGPT and Vicuna-13B in the Vicun A Benchmark’s question prompt. |
| Outcome: | The proposed framework alleviates evaluation bias, resulting in closer alignment with human judgments. |
ReCon: Active Defense against Large Vision-Language Model Jailbreaks via Reverse Safety Concept Injection (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing defense strategies neglect visual threats and lack of fine-grained specificity regarding specific attack semantics. |
| Approach: | They propose a black-box defense framework that maps unsafe concepts to fine-grained, constructive Safe Concepts. |
| Outcome: | a new black-box defense framework enhances robustness against jailbreak attacks . it maps detected unsafe concepts to fine-grained, constructive Safe Concepts . the proposed framework is available for free at http://www.epa.org/recon/ . |
Graph-Structured Speculative Decoding (2024.findings-acl)
Copied to clipboard
Zhuocheng Gong, Jiahao Liu, Ziyue Wang, Pengfei Wu, Jingang Wang, Xunliang Cai, Dongyan Zhao, Rui Yan
| Challenge: | Speculative decoding is a promising technique to accelerate the inference of Large Language Models. |
| Approach: | They propose a method that uses a token graph to record multiple sequence hypotheses within a single draft stage. |
| Outcome: | The proposed method significantly accelerates the inference of Large Language Models (LLMs) it allows the LLM to choose from and select the longest sequence that meets its standards. |
Learning from Contrasts: Synthesizing Reasoning Paths from Diverse Search Trajectories (2026.acl-long)
Copied to clipboard
| Challenge: | MCTS methods retain only the single highest-reward trajectory, discarding comparative signals present in the many explored paths. |
| Approach: | They propose a framework that transforms supervision extraction into a synthesis procedure. |
| Outcome: | The proposed framework matches or exceeds baselines on 60K CRPS-synthesized examples on out-of-domain benchmarks. |
Omni-Chart-600K: A Comprehensive Dataset of Chart Types for Chart Understanding (2025.findings-naacl)
Copied to clipboard
Shulei Wang, Shuai Yang, Wang Lin, Zirun Guo, Sihang Cai, Hai Huang, Ye Wang, Jingyuan Chen, Tao Jin
| Challenge: | Existing chart-related training methods lack capabilities in information extraction, mathematical reasoning, and understanding of multiple chart types. |
| Approach: | They propose a two-stage training strategy and method for jointly training a vision encoder tailored for multi-type charts to address the deficiencies in chart types and limited scope of chart tasks in existing datasets. |
| Outcome: | The proposed dataset includes 21 diverse chart types and tasks, including data retrieval and mathematical reasoning. |
Task-oriented Domain-specific Meta-Embedding for Text Classification (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods neglect domain-specific knowledge and use the same word embedding for each word in all domain-specified datasets. |
| Approach: | They propose a method to incorporate domain-specific and task-oriented information into meta-embeddings by combining pre-trained word embeddings. |
| Outcome: | The proposed method performs well on four text classification datasets and shows that it is compatible with existing methods. |
Distributional Clarity: The Hidden Driver of RL-Friendliness in Large Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | RL-friendly models exhibit intra-class compactness and inter-class separation in probability assignments . under identical training, Qwen models achieve substantial gains, while others like Llama yield limited improvements. |
| Approach: | They propose a method to quantify distributional clarity in probability space . they show distributional clearness is a trainable property underlying RL-Friendliness . |
| Outcome: | The proposed model families achieve substantial gains under identical training, while others like Llama yield limited improvements. |
ChartVerse: Scaling Chart Reasoning via Reliable Programmatic Synthesis from Scratch (2026.acl-long)
Copied to clipboard
Zheng Liu, Honglin Lin, Xiaoyang Wang, Xin Gao, Yu Li, Mengzhang Cai, Yun Zhu, Zhanping Zhong, Qizhi Pei, Zhuoshi Pan, Xiaoran Shang, Conghui He, Bin Cui, Wentao Zhang, Lijun Wu
| Challenge: | Existing open-source vision language models lack high-quality training data for chart reasoning . current models are simplistic and repetitive, while associated QA pairs are prone to hallucinations . |
| Approach: | They propose a framework to synthesize complex charts and reliable reasoning data from scratch. |
| Outcome: | Experimental results show that ChartVerse-8B surpasses existing models in QA and difficulty . lack of high-quality training data hampers development of open-source models . |
Retrieval-guided Dialogue Response Generation via a Matching-to-Generation Framework (D19-1)
Copied to clipboard
| Challenge: | generative models for end-to-end sequence generation have been shown promising for this task . however, how to precisely extract a skeleton and how to effectively train a retrieval-guided response generator is still challenging. |
| Approach: | They propose a framework where skeleton extraction is made by an interpretable matching model and a retrieval-guided response generator is followed by a separate generator. |
| Outcome: | The proposed framework outperforms baseline models in a variety of experiments. |
Generating User-Engaging News Headlines (2023.acl-long)
Copied to clipboard
| Challenge: | Personalized news recommendation systems present the same headline to all users, making it difficult for them to understand the connection between their interests and the recommended article. |
| Approach: | They propose a framework that incorporates user profiling to generate personalized headlines and a combination of automated and human evaluation methods to determine user preference for personalized headline generation. |
| Outcome: | The proposed framework can generate personalized headlines that meet the needs of a diverse audience. |
Beyond Code: Evaluate Thought Steps for Complex Code Generation (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing efforts to generate code in C++ rely on relatively simple programming problems . large language models (LLMs) pre-trained on numerous code data have opened up new opportunities for code generation. |
| Approach: | They propose a task that evaluates the quality of thought steps and code implementation . they construct a dataset of complex programming problems in C++ . |
| Outcome: | The proposed task evaluates the quality of thought steps and code implementation in a C++ programming language. |
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. |
Translating a Math Word Problem to a Expression Tree (D18-1)
Copied to clipboard
| Challenge: | Sequence-to-sequence (SEQ2SEQ) models have been successfully applied to automatic math word problem solving. |
| Approach: | They propose an equation normalization method to normalize duplicated equations and propose an ensemble model to combine their advantages. |
| Outcome: | The proposed model outperforms the previous state-of-the-art models on the math word problem solving. |
Towards Medical Complex Reasoning with LLMs through Medical Verifiable Problems (2025.findings-acl)
Copied to clipboard
| Challenge: | OpenAI o1 has been a significant milestone in large language model development . however, most research in reasoning has focused on mathematical tasks . medical domains require robust reasoning to provide reliable answers . |
| Approach: | They propose a method to verify medical reasoning using a medical verifier . they also propose RL and reinforcement learning to enhance reasoning . |
| Outcome: | The proposed method outperforms general and medical-specific baselines using only 40K verifiable problems. |
Query2Triple: Unified Query Encoding for Answering Diverse Complex Queries over Knowledge Graphs (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Complex Query Answering (CQA) is a challenge task of Knowledge Graphs due to incompleteness of KGs. |
| Approach: | They propose a query embedding approach that decouples the training for simple and complex queries. |
| Outcome: | The proposed approach decouples training for simple and complex queries and achieves state-of-the-art performance over three public benchmarks. |
Understanding GUI Agent Localization Biases through Logit Sharpness (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Multimodal large language models often exhibit hallucinations that compromise reliability . despite promising performance, these models often display systematic localization errors . |
| Approach: | They propose a framework that categorizes model predictions into four distinct types . they propose metric that evaluates alignment between semantic continuity and logits distribution . |
| Outcome: | The proposed framework categorizes model predictions into four different types . it reveals nuanced failure modes beyond traditional accuracy metrics . |
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. |
A Thorough Examination of Decoding Methods in the Era of LLMs (2024.emnlp-main)
Copied to clipboard
| Challenge: | Decoding methods are essential for converting language models from next-token predictors into practical task solvers. |
| Approach: | They propose to evaluate decoding methods in general-purpose large language models . they find that decoding method performance is notably task-dependent . |
| Outcome: | The proposed methods perform task-dependently and are influenced by alignment, model size, and quantization. |
Knowledge-Empowered Representation Learning for Chinese Medical Reading Comprehension: Task, Model and Resources (2021.findings-acl)
Copied to clipboard
| Challenge: | MRC is a popular task in NLP, aiming to understand a passage and answer the relevant questions. |
| Approach: | They propose a multi-target machine learning task for the medical domain that predicts answers to medical questions and corresponding support sentences from medical information sources simultaneously. |
| Outcome: | The proposed model outperforms baselines by fusing context-aware and knowledge-awful token representations. |
AMO-Bench: Large Language Models Still Struggle in High School Math Competitions (2026.findings-acl)
Copied to clipboard
Junlin Liu, Shengnan An, Shuang Zhou, Dan Ma, Yehao Lin, Xinxuan Lv, Xuanlin Wang, Xiaoyu Li, Ziwen Wang, Xuezhi Cao, Xunliang Cai
| Challenge: | Existing benchmarks for mathematical reasoning are becoming less effective due to performance saturation. |
| Approach: | They propose to use a mathematical reasoning benchmark with Olympiad difficulty to evaluate top-tier LLMs. |
| Outcome: | The proposed benchmarks are cross-validated by experts to meet IMO difficulty standards and entirely original problems to prevent performance leakages from data memorization. |
Self-Renewal Prompt Optimizing with Implicit Reasoning (2024.findings-emnlp)
Copied to clipboard
Zihan Liang, Ben Chen, Zhuoran Ran, Zihan Wang, Huangyu Dai, Yufei Ma, Dehong Gao, Xiaoyan Cai, Libin Yang
| Challenge: | Recent advances in NLP have been driven by the development of Large Language Models (LLMs). |
| Approach: | They propose a self-renewal approach to optimize LLM outputs to better align with human preferences without supervised fine-tuning. |
| Outcome: | The proposed approach improves outputs to better align with human preferences across LLMs and tasks without supervised fine-tuning. |
Bridging the Pose-Semantic Gap: A Cascade Framework for Text-Based Person Anomaly Search (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for person anomaly search fail to address the complexities of real-world security, authors say . Existing approaches fail to detect subtle semantic distinctions, authors argue . |
| Approach: | They propose a framework that decouples retrieval into two stages . structure-aware coarse retrieval and detective squad interaction are proposed . |
| Outcome: | The proposed framework achieves state-of-the-art performance by balancing efficiency and semantic reasoning. |
Enhancing Reasoning Abilities of Small LLMs with Cognitive Alignment (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods to distill chain-of-thought (CoT) results from large language reasoning models (LRMs) to small models are ineffective and require substantial amount of annotated data. |
| Approach: | They propose a Critique-Rethink-Verify system for training small language reasoning models that can be critiquized according to the cognitive capabilities of smaller models. |
| Outcome: | The proposed system outperforms other methods on challenging reasoning benchmarks. |
Exploiting Emotion-Semantic Correlations for Empathetic Response Generation (2023.findings-emnlp)
Copied to clipboard
Zhou Yang, Zhaochun Ren, Wang Yufeng, Xiaofei Zhu, Zhihao Chen, Tiecheng Cai, Wu Yunbing, Yisong Su, Sibo Ju, Xiangwen Liao
| Challenge: | Empathetic response generation aims to generate empathetic responses by understanding the speaker’s emotional feelings from the language of dialogue. |
| Approach: | They propose a dynamical Emotion-Semantic Correlation Model (ESCM) which constructs dynamic emotion-semantics through the interaction of context and emotions. |
| Outcome: | The proposed model understands emotions more accurately and expresses fluent and informative empathetic responses. |
A Survey of Confidence Estimation and Calibration in Large Language Models (2024.naacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have demonstrated impressive capabilities across a wide range of tasks in various domains, but they can be unreliable due to factual errors in their generations. |
| Approach: | They summarize recent advances in LLM confidence estimation and calibration and outline their main lessons learned. |
| Outcome: | The proposed methods can be used to assess the reliability of models and to calibrate them across tasks. |
Consolidation or Adaptation? PRISM: Disentangling SFT and RL Data via Gradient Concentration (2026.acl-long)
Copied to clipboard
Yang Zhao, Yangou Ouyang, Xiao Ding, Hepeng Wang, Bibo Cai, Kai Xiong, Jinglong Gao, Zhouhao Sun, Li Du, Bing Qin, Ting Liu
| Challenge: | Existing data arbitration strategies for large language model training rely on surface-level heuristics that fail to diagnose intrinsic learning needs. |
| Approach: | They propose a framework that arbitrates data based on its degree of cognitive conflict with the model's existing knowledge. |
| Outcome: | Extensive experiments on WebShop and ALFWorld show that PRISM outperforms state-of-the-art hybrid methods while reducing computational costs by up to 3.22 . |
Primacy Effect of ChatGPT (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing machine learning models may lead to poor performance in discriminative natural language understanding tasks. |
| Approach: | They propose to use ChatGPT to query large amounts of human-written text to find the answer to a question. |
| Outcome: | The proposed model has a high chance to select labels at earlier positions as the answer. |
Multimodal Causal Reasoning Benchmark: Challenging Multimodal Large Language Models to Discern Causal Links Across Modalities (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing MLLMs lack robustness in multimodal causal reasoning compared to their performance in textual settings. |
| Approach: | They propose a novel multimodal chain-of-thought (CoT) reasoning benchmark that leverages siamese images and text pairs to challenge MLLMs. |
| Outcome: | The proposed benchmark leverages siamese images and text pairs to challenge MLLMs. |
Towards General Agentic Intelligence via Environment Scaling (2026.findings-acl)
Copied to clipboard
Runnan Fang, Shihao Cai, Baixuan Li, Jialong Wu, Guangyu Li, Wenbiao Yin, Xinyu Wang, Xiaobin Wang, Liangcai Su, Zhen Zhang, Shibin Wu, Zhengwei Tao, Yong Jiang, Pengjun Xie, Ningyu Zhang, Fei Huang, Wentao Zhang, Jingren Zhou
| Challenge: | Diverse real-world APIs require precise, robust function-calling intelligence, which needs agents to develop these capabilities through interaction in varied environments. |
| Approach: | They propose a framework that scales up environments to enable agentic intelligence . they use a two-phase agent fine-tuning strategy to first endow agents with basic agentic capabilities, then specializing them for domain-specific contexts. |
| Outcome: | Experiments on -bench, -Bench, and ACEBench show that the model significantly enhances the models’ function-calling capability. |
Improving Input-label Mapping with Demonstration Replay for In-context Learning (2023.findings-emnlp)
Copied to clipboard
| Challenge: | In-context learning (ICL) is an emerging capability of large autoregressive language models where a few demonstrations are appended to the input to enhance the model’s understanding of downstream NLP tasks without directly adjusting the model parameters. |
| Approach: | They propose a method where a few demonstrations are appended to the input to enhance the model's understanding of downstream NLP tasks without directly adjusting the model parameters. |
| Outcome: | The proposed method significantly improves the input-label mapping in ICL demonstrations. |
SciAssess: Benchmarking LLM Proficiency in Scientific Literature Analysis (2025.findings-naacl)
Copied to clipboard
Hengxing Cai, Xiaochen Cai, Junhan Chang, Sihang Li, Lin Yao, Wang Changxin, Zhifeng Gao, Hongshuai Wang, Li Yongge, Mujie Lin, Shuwen Yang, Jiankun Wang, Mingjun Xu, Jin Huang, Xi Fang, Jiaxi Zhuang, Yuqi Yin, Yaqi Li, Changhong Chen, Zheng Cheng, Zifeng Zhao, Linfeng Zhang, Guolin Ke
| Challenge: | Existing benchmarks fail to adequately evaluate the proficiency of Large Language Models (LLMs) Existing standards do not cover the skills needed to evaluate LLMs in scientific literature analysis. |
| Approach: | They propose a benchmark to evaluate the proficiency of large language models in scientific literature analysis. |
| Outcome: | SciAssess evaluates 11 LLMs on multiple tasks across scientific fields. |
Improving Empathetic Dialogue Generation by Dynamically Infusing Commonsense Knowledge (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing work on generating empathetic responses by utilizing the speaker's emotion has not been successful. |
| Approach: | They propose an approach which incorporates an adaptive module for commonsense knowledge selection to ensure consistency between the generated empathetic responses and the speaker’s situation. |
| Outcome: | The proposed approach outperforms baseline models in both automatic and human evaluations, exhibiting the generation of more coherent and empathetic responses. |
Towards Injecting Medical Visual Knowledge into Multimodal LLMs at Scale (2024.emnlp-main)
Copied to clipboard
Junying Chen, Chi Gui, Ruyi Ouyang, Anningzhe Gao, Shunian Chen, Guiming Chen, Xidong Wang, Zhenyang Cai, Ke Ji, Xiang Wan, Benyou Wang
| Challenge: | Multimodal large language models (MLLMs) lack visual knowledge in medical applications due to data privacy concerns and high annotation costs. |
| Approach: | They refined medical image-text pairs from PubMed and employed MLLMs (GPT-4V) to denoise and reformat the data. |
| Outcome: | The proposed model significantly improves the MMMU Health & Medicine track and shows that it can be used in multimodal scenarios. |
Large-Scale Diverse Synthesis for Mid-Training (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing data synthesis methods generate simplistic and homogeneous QA pairs with limited scale and diversity. |
| Approach: | They propose a framework to synthesize large-scale, diverse, and high-quality QA data for mid-training. |
| Outcome: | The proposed framework improves on 500B-token BoostQA data over pre-training benchmarks. |
Unleashing Large Language Models’ Proficiency in Zero-shot Essay Scoring (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Recent advances in automated essay scoring (AES) have relied on labeled essays, requiring tremendous cost and expertise for their acquisition. |
| Approach: | They propose a zero-shot prompting framework that automatically decomposes writing proficiency into distinct traits and generates scoring criteria for each trait. |
| Outcome: | The proposed framework outperforms straightforward prompting (Vanilla) on TOEFL11 and ASAP, while the small-sized Llama2-13b-chat significantly outperformed ChatGPT. |
LLM Critics Help Catch Bugs in Mathematics: Towards a Better Mathematical Verifier with Natural Language Feedback (2025.findings-acl)
Copied to clipboard
Bofei Gao, Zefan Cai, Runxin Xu, Peiyi Wang, Ce Zheng, Runji Lin, Keming Lu, Dayiheng Liu, Chang Zhou, Wen Xiao, Tianyu Liu, Baobao Chang
| Challenge: | Existing mathematical verifiers are trained with binary classification labels, which are not informative enough for the model to accurately assess the solutions. |
| Approach: | They propose a natural language feedback-enhanced verifier that can validate the correctness of response generated by policy models by constructing automatically generated training data and a two-stage training paradigm. |
| Outcome: | The proposed verifier significantly improves in verification and reinforcement learning and alleviates data-demanding problems of the reward model. |
VistaWise: Building Cost-Effective Agent with Cross-Modal Knowledge Graph for Minecraft (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) have shown significant promise in embodied decision-making tasks within virtual open-world environments, but lack domain-specific knowledge. |
| Approach: | They propose a cost-effective agent framework that integrates cross-modal domain knowledge and finetunes a dedicated object detection model for visual analysis. |
| Outcome: | The proposed framework reduces the requirement for domain-specific training data from millions of samples to a few hundred. |
Empowering Large Language Model for Continual Video Question Answering with Collaborative Prompting (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing VideoQA models struggle to adapt to new questions or tasks posed by newly available content. |
| Approach: | They propose a continual learning framework that fine-tunes a large language model for a sequence of tasks and integrates specific question constraint prompting, knowledge acquisition prompting and visual temporal awareness prompting. |
| Outcome: | The proposed model achieves 55.14% accuracy on both NExT-QA and DramaQA datasets and 71.24% accuracy for DramaQA. |
HiEdit: Lifelong Model Editing with Hierarchical Reinforcement Learning (2026.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to lifelong model editing apply parameter perturbations to static and dense layers for all instances. |
| Approach: | They propose a hierarchical reinforcement learning framework that identifies the most knowledge-relevant layers for each editing instance. |
| Outcome: | The proposed framework boosts the performance of the competitive RLEdit by 8.48% with perturbing only half of the layers per edit. |
EmpCRL: Controllable Empathetic Response Generation via In-Context Commonsense Reasoning and Reinforcement Learning (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing studies lack the perception of fine-grained dialogue emotion propagation, and have limitations in reasoning about the intentions of users on cognition, which affect the quality of empathetic response. |
| Approach: | They propose to use commonsense reasoning and reinforcement learning to generate empathetic response based on in-context commonsensing and contextual reasoning to broaden cognitive boundaries. |
| Outcome: | The proposed model outperforms state-of-the-art models in automatic and human evaluation. |
Multi-Programming Language Sandbox for LLMs (2025.acl-demo)
Copied to clipboard
Shihan Dou, Jiazheng Zhang, Jianxiang Zang, Yunbo Tao, Weikang Zhou, Haoxiang Jia, Shichun Liu, Yuming Yang, Shenxi Wu, Zhiheng Xi, Muling Wu, Rui Zheng, Changze Lv, Limao Xiong, Shaoqing Zhang, Lin Zhang, Wenyu Zhan, Rongxiang Weng, Jingang Wang, Xunliang Cai, Yueming Wu, Ming Wen, Yixin Cao, Tao Gui, Xipeng Qiu, Qi Zhang, Xuanjing Huang
| Challenge: | MPLSandbox is an out-of-the-box multi-programming language sandbox designed to provide unified and comprehensive feedback from compiler and analysis tools for Large Language Models (LLMs). |
| Approach: | They propose a multi-programming language sandbox that provides unified feedback from compilers and analysis tools for Large Language Models. |
| Outcome: | The proposed multi-language sandbox can provide comprehensive feedback from compilers and analysis tools for large language models (LLMs). |
Efficient Learned Data Compression via Dual-Stream Feature Decoupling (2026.acl-long)
Copied to clipboard
| Challenge: | Learned data compression has achieved superior compression ratios, but balancing precise probability modeling with system efficiency remains challenging. |
| Approach: | They propose a Dual-Stream Multi-Scale Decoupler that disentangles local and global contexts to replace deep serial processing with shallow parallel streams. |
| Outcome: | The proposed method achieves state-of-the-art performance in both compression ratio and throughput while maintaining the lowest latency and memory usage. |
RTADev: Intention Aligned Multi-Agent Framework for Software Development (2025.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are efficient assistants to humans in software development tasks, but they can cause errors during the development process. |
| Approach: | They propose an intention aligned multi-agent framework that ensures that all agents work based on a consensus. |
| Outcome: | The proposed framework reduces errors and improves the quality of generated software code. |
LinkQA: Synthesizing Diverse QA from Multiple Seeds Strongly Linked by Knowledge Points (2026.acl-long)
Copied to clipboard
| Challenge: | Existing training data is limited in high-quality training data, limiting the ability to produce high-performance LLMs. |
| Approach: | They propose a KP-graph-based synthesis framework that extracts KPs from QA seed data and constructs a graph of KP data from multiple seeds strongly linked by KP. |
| Outcome: | The proposed framework enables flexible control over discipline and difficulty distributions while balancing KP coverage and popularity. |
FineSteer: A Unified Framework for Fine-Grained Inference-Time Steering in Large Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for inference-time steering fail to be effective, utility-preserving and training-efficient due to rigid, one-size-fits-all designs and limited adaptability. |
| Approach: | They propose a steering framework that decomposes inference-time steering into two stages . they propose 'conditional steering' mechanism that preserves model utility by avoiding unnecessary steering . a 'mixture-of-Steering-Experts' mechanism captures multimodal nature of desired steering behaviors . |
| Outcome: | The proposed framework outperforms the state-of-the-art methods on safety and truthfulness benchmarks. |
FRAME: Boosting LLMs with A Four-Quadrant Multi-Stage Pretraining Strategy (2025.findings-acl)
Copied to clipboard
Xuemiao Zhang, Feiyu Duan, Xu Liangyu, Yongwei Zhou, Sirui Wang, Rongxiang Weng, Jingang Wang, Xunliang Cai
| Challenge: | Multi-stage pretraining methods lack quantitative criteria for data partitioning and instead rely on intuitive heuristics. |
| Approach: | They propose a Four-quadRAnt Multi-stage prEtraining strategy that partitions data into four quadrants to achieve significant loss reductions four times. |
| Outcome: | The proposed strategy achieves 16.8% improvement over random across MMLU and CMMLU for the 3B model. |
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. |
DoTAT: A Domain-oriented Text Annotation Tool (2022.acl-demo)
Copied to clipboard
| Challenge: | DoTAT is a domain-oriented text annotation tool that can reduce the time for event annotation by 19.7% . the tool supports multi-person collaborative process with automatically merging and review . |
| Approach: | They propose a domain-oriented text annotation tool called DoTAT . it provides multi-person collaborative process with automatic merging and review . |
| Outcome: | The proposed tool can reduce the time for event annotation by 19.7% compared with existing tools. |
Iterative Self-Correction for Text-Driven Person Re-Identification with Large Vision-Language Models (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for Person Re-Identification (ReID) adopt a static "one-pass" paradigm, converting images to text once for retrieval. |
| Approach: | They propose a framework that reformulates ReID as an iterative "Think-and-Refine" process. |
| Outcome: | The proposed framework outperforms state-of-the-art methods in complex occlusion scenarios. |
Low-code LLM: Graphical User Interface over Large Language Models (2024.naacl-demo)
Copied to clipboard
Yuzhe Cai, Shaoguang Mao, Wenshan Wu, Zehua Wang, Yaobo Liang, Tao Ge, Chenfei Wu, WangYou WangYou, Ting Song, Yan Xia, Nan Duan, Furu Wei
| Challenge: | Low-code LLM is a visual programming interface that allows users to incorporate their ideas into the process without writing trivial prompts. |
| Approach: | They propose a human-LLM interaction framework that incorporates low-code visual programming interactions to achieve more controllable and stable responses. |
| Outcome: | The proposed framework enables users to incorporate ideas into the process without writing trivial prompts. |
Leveraging Unpaired Feedback for Long-Term LLM-based Recommendation Tuning (2025.findings-emnlp)
Copied to clipboard
| Challenge: | a recent study highlights unpaired feedback as a key challenge for long-term LLM-based recommenders . unpaired user feedback is crucial for improving LLMs in dynamic user environments, authors say . |
| Approach: | They propose a framework that incorporates unpaired feedback into LLMs to improve long-term recommendation performance. |
| Outcome: | The proposed framework improves long-term recommendation performance by incorporating unpaired feedback without requiring paired supervision. |
Not All Contexts Are Equal: Teaching LLMs Credibility-aware Generation (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing RAG paradigms suffer from the impact of flawed information introduced during the retrieval phrase, thereby diminishing the reliability and correctness of the generated output. |
| Approach: | They propose a framework that empowers models to discern and process information based on its credibility. |
| Outcome: | The proposed framework outperforms existing models with retrieval augmentation and exhibits robustness despite increasing noise in the context. |
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. |
PA-RAG: RAG Alignment via Multi-Perspective Preference Optimization (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing approaches to optimize RAG generators fail to align with RAG requirements thoroughly. |
| Approach: | They propose a method for optimizing the RAG generator from multiple preference perspectives to align with RAG requirements comprehensively. |
| Outcome: | The proposed method improves the performance of RAG generators by incorporating retrieved documents into the prompt. |
Should We Rely on Entity Mentions for Relation Extraction? Debiasing Relation Extraction with Counterfactual Analysis (2022.naacl-main)
Copied to clipboard
Yiwei Wang, Muhao Chen, Wenxuan Zhou, Yujun Cai, Yuxuan Liang, Dayiheng Liu, Baosong Yang, Juncheng Liu, Bryan Hooi
| Challenge: | Existing studies rely on entity information for sentence-level relation extraction (RE) but this can leak superficial and spurious clues of relations. |
| Approach: | They propose to use entity mentions to extract relations from textual context . they use a causal graph to model dependencies between variables in RE models . |
| Outcome: | The proposed method yields significant gains on both effectiveness and generalization for RE. |
How Do Your Code LLMs perform? Empowering Code Instruction Tuning with Really Good Data (2024.emnlp-main)
Copied to clipboard
Yejie Wang, Keqing He, Dayuan Fu, Zhuoma GongQue, Heyang Xu, Yanxu Chen, Zhexu Wang, Yujia Fu, Guanting Dong, Muxi Diao, Jingang Wang, Mengdi Zhang, Xunliang Cai, Weiran Xu
| Challenge: | Recent research has shown that code pre-trained models improve coding capabilities. |
| Approach: | They propose a code data pruning strategy to identify which datasets are high-quality code instruction data. |
| Outcome: | The proposed model achieves state-of-the-art performance using fewer training data. |
Deep Cognitive Reasoning Network for Multi-hop Question Answering over Knowledge Graphs (2021.findings-acl)
Copied to clipboard
| Challenge: | Knowledge Graphs (KGs) store structured human knowledge with nodes and edges being entities and relations between them. |
| Approach: | They propose a deep cognitive reasoning network that uses two phases to find answers in large candidate entity sets. |
| Outcome: | The proposed method significantly outperforms state-of-the-art methods on benchmark datasets. |
MRFD: Multi-Region Fusion Decoding with Self-Consistency for Mitigating Hallucinations in LVLMs (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Large Vision-Language Models often produce hallucinations due to the limited ability to verify information in different regions of the image. |
| Approach: | a new decoding method improves factual grounding by modeling inter-region consistency . the method identifies salient regions using cross-attention and generates initial responses for each . |
| Outcome: | a training-free decoding method reduces hallucinations and improves response consistency . the proposed method generates initial responses for each region and weights reliability weights among responses . |
MoLAN: A Unified Modality-Aware Noise Dynamic Editing Framework for Multimodal Sentiment Analysis (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to multimodal sentiment analysis treat entire modality as an independent unit for feature enhancement or denoising, which often suppresses redundant noise at the cost of weakening critical information. |
| Approach: | They propose a ModaLity-aware noise dynAmic editiNg framework that performs modality-awful block partitioning by dividing features of each modality into multiple blocks. |
| Outcome: | Experiments on five models and four datasets show that MoLAN+ achieves the state-of-the-art performance. |
SemVink: Advancing VLMs’ Semantic Understanding of Optical Illusions via Visual Global Thinking (2025.emnlp-main)
Copied to clipboard
| Challenge: | Vision-language models excel in semantic tasks but fail at detecting hidden content . current architectures prioritize abstract reasoning over low-level visual operations . |
| Approach: | They propose a benchmark to test vision-language models that can detect hidden content . they propose HC-Bench to scale images to low resolutions to unlock 99% accuracy . |
| Outcome: | HC-Bench shows that leading VLMs achieve near-zero accuracy even with explicit prompting . et al.: current models prioritize abstract reasoning over low-level visual operations . they urge a shift toward hybrid models bridging gap between computational vision and human cognition . |
AdaSwitch: Adaptive Switching between Small and Large Agents for Effective Cloud-Local Collaborative Learning (2024.emnlp-main)
Copied to clipboard
Hao Sun, Jiayi Wu, Hengyi Cai, Xiaochi Wei, Yue Feng, Bo Wang, Shuaiqiang Wang, Yan Zhang, Dawei Yin
| Challenge: | Recent advances in large language models have been remarkable . users face a choice between using cloud-based LLMs for generation quality or local-based ones for lower computational cost . |
| Approach: | They propose a new LLM utilization paradigm that facilitates collaborative operation . they evaluate AdaSwitch across 7 benchmarks and compare it to other LLMs . |
| Outcome: | The proposed model improves performance of local and cloud agents across 7 benchmarks . it achieves competitive results compared to the cloud agent while utilizing less computational overhead. |
Enhancing Advanced Visual Reasoning Ability of Large Language Models (2024.emnlp-main)
Copied to clipboard
| Challenge: | Recent advances in Vision-Language (VL) research have sparked new benchmarks for complex visual reasoning, challenging models’ advanced reasoning ability. |
| Approach: | They propose a novel multi-modal in-context learning methodology to enhance LLMs’ contextual understanding and reasoning. |
| Outcome: | The proposed model achieves SOTA performance among all visual reasoning tasks and achieves a 'higher level of accuracy' than previous models. |
ReMamba: Equip Mamba with Effective Long-Sequence Modeling (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Mamba models demonstrate superior inference efficiency and competitive performance on short-context tasks, but their capacity to comprehend long contexts is limited compared to transformer-based models. |
| Approach: | They propose a model which incorporates selective compression and adaptation techniques within a two-stage re-forward process, incurring minimal additional inference costs overhead. |
| Outcome: | The proposed model improves on the LongBench and L-Eval benchmarks by 3.2 and 1.6 points and attains performance almost on par with same-size transformer models. |
Attribution-Based Analysis and Optimization of Modular Agentic Workflows (2026.findings-acl)
Copied to clipboard
Yingxuan Yang, Bo Huang, Siyuan Qi, Chao Feng, Haoyi Hu, Yuxuan Zhu, Jinbo Hu, Haoran Zhao, Ziyi He, Xiao Liu, ZongYu Wang, Muning Wen, Lin Qiu, Xuezhi Cao, Xunliang Cai, Yong Yu, Weinan Zhang
| Challenge: | Large Language Models (LLMs) have driven the rise of agentic workflows . yet, how can we attribute performance gains to individual upgrades and their interactions? |
| Approach: | They propose a game-theoretic framework that models component upgrades as players and evaluates component coalitions to compute Shapley values. |
| Outcome: | The proposed framework provides interaction-aware attribution and recommendation for model allocation under a fixed workflow structure. |
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. |
FaithLM: Towards Faithful Explanations for Large Language Models (2026.eacl-long)
Copied to clipboard
Yu-Neng Chuang, Guanchu Wang, Chia-Yuan Chang, Ruixiang Tang, Shaochen Zhong, Fan Yang, Andrew Wen, Mengnan Du, Xuanting Cai, Vladimir Braverman, Xia Hu
| Challenge: | Large language models (LLMs) produce natural language explanations, but they lack faithfulness and do not reflect the evidence the model uses to decide. |
| Approach: | They propose a model-agnostic framework that evaluates and improves the faithfulness of LLM explanations without token masking or task-specific heuristics. |
| Outcome: | The proposed framework improves faithfulness of large language models without masking or heuristics. |
DRK: Discriminative Rule-based Knowledge for Relieving Prediction Confusions in Few-shot Relation Extraction (2022.coling-1)
Copied to clipboard
| Challenge: | Existing methods to identify relation type in low-resource scenario fall into prediction confusions owing to the limited inference ability over shallow text features. |
| Approach: | They propose a discriminative rule-based knowledge method to identify the relation type between entities in a given text in the low-resource scenario. |
| Outcome: | The proposed method improves on four types of meta tasks with a 6.0% accuracy gain on average. |
MAESTRO: Meta-learning Adaptive Estimation of Scalarization Trade-offs for Reward Optimization (2026.acl-long)
Copied to clipboard
Yang Zhao, Hepeng Wang, Xiao Ding, Yangou Ouyang, Bibo Cai, Kai Xiong, Jinglong Gao, Zhouhao Sun, Li Du, Bing Qin, Ting Liu
| Challenge: | Group-Relative Policy Optimization (GRPO) has emerged as an efficient paradigm for aligning Large Language Models (LLMs), but its efficacy is confined to domains with verifiable ground truths. |
| Approach: | They propose a meta-cognitive orchestration layer that treats reward scalarization as a dynamic latent policy, leveraging the model’s terminal hidden states as 'a semantic bottleneck' . Across seven benchmarks, MAESTRO consistently outperforms single-reward and static multi-objective baselines while preserving the efficiency advantages of GRPO. |
| Outcome: | The proposed model outperforms single-reward and static multi-objective baselines while preserving efficiency advantages. |
Revisiting Automated Evaluation for Long-form Table Question Answering (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing automated metrics for long-form table question answering (LFTQA) are poorly correlated with human judgments and fail to distinguish between factually accurate responses and those that are factual incorrect. |
| Approach: | They propose to use a meta-evaluation dataset to assess the effectiveness of LLM-based LFTQA systems. |
| Outcome: | The proposed meta-evaluation dataset includes 2,988 human-annotated examples. |
MoDULA: Mixture of Domain-Specific and Universal LoRA for Multi-Task Learning (2024.emnlp-main)
Copied to clipboard
Yufei Ma, Zihan Liang, Huangyu Dai, Ben Chen, Dehong Gao, Zhuoran Ran, Wang Zihan, Linbo Jin, Wen Jiang, Guannan Zhang, Xiaoyan Cai, Libin Yang
| Challenge: | Recent advances in open-source Large Language Models (LLMs) have achieved notable successes in natural language processing. |
| Approach: | They propose a Parameter Efficient Fine-Tuning paradigm for improved fine-tuning and parameter efficiency in multi-task learning. |
| Outcome: | The proposed model outperforms existing methods on multi-task learning while reducing training costs by over 80% without losing general capability. |
NeuroSym-Cal: Bridging the Reasoning-Execution Gap in Code Generation via Hierarchical Calibration (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing calibration methods rely on the assumption that consensus implies correctness . Existing methods fail under systematic errors, leading to miscalibrated high-confidence predictions. |
| Approach: | They propose a hierarchical calibration framework that measures confidence at two levels . they propose sensitivity analysis to measure local curvature of deductive process . |
| Outcome: | The proposed framework de-saturates overconfident errors and improves selective generation performance on OOD benchmarks. |
Guiding Abstractive Dialogue Summarization with Content Planning (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for abstractive dialogue summarization struggle to maintain factual consistency between dialogue and summary. |
| Approach: | They propose a coarse-to-fine model for generating abstractive dialogue summaries and introduce a fact-aware reinforcement learning objective that improves the fact consistency between the dialogue and the generated summary. |
| Outcome: | The proposed model improves the quality of the generated summary, especially in coherence and consistency. |
Advancing Fine-Grained Visual Understanding with Multi-Scale Alignment in Multi-Modal Models (2025.emnlp-main)
Copied to clipboard
Wei Wang, Zhaowei Li, Qi Xu, Linfeng Li, YiQing Cai, Botian Jiang, Hang Song, Xingcan Hu, Pengyu Wang, Li Xiao
| Challenge: | Recent advances in multi-modal large language models have demonstrated remarkable capabilities in multimodal understanding, reasoning, and interaction. |
| Approach: | They propose a method that effectively aligns and integrates multi-scale knowledge of objects . they use a pipeline that provides over 300K essential training data to enhance alignment . |
| Outcome: | The proposed method effectively aligns and integrates multi-scale knowledge of objects, including texts, coordinates, and images. |
DolphCoder: Echo-Locating Code Large Language Models with Diverse and Multi-Objective Instruction Tuning (2024.acl-long)
Copied to clipboard
Yejie Wang, Keqing He, Guanting Dong, Pei Wang, Weihao Zeng, Muxi Diao, Weiran Xu, Jingang Wang, Mengdi Zhang, Xunliang Cai
| Challenge: | Numerous code large language models (LLMs) have been proposed to enhance code generation performance. |
| Approach: | They propose a diverse instruction model DolphCoder with self-evaluating for code generation that learns diverse instruction targets and combines a code evaluation objective to enhance its code generation ability. |
| Outcome: | The proposed model achieves superior performance on the HumanEval and MBPP benchmarks, demonstrating new insights for future code instruction tuning work. |
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). |
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. |
GroundingGPT: Language Enhanced Multi-modal Grounding Model (2024.acl-long)
Copied to clipboard
Zhaowei Li, Qi Xu, Dong Zhang, Hang Song, YiQing Cai, Qi Qi, Ran Zhou, Junting Pan, Zefeng Li, Vu Tu, Zhida Huang, Tao Wang
| Challenge: | Existing multi-modal large language models focus on capturing global information while neglecting the fine-grained local information in multimodal inputs. |
| Approach: | They propose an end-to-end language enhanced multi-modal grounding model that performs fine-grained grounding tasks for image, video and audio. |
| Outcome: | The proposed model achieves impressive fine-grained understanding of multi-modal inputs while maintaining or improving its global comprehension capabilities. |
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. |
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. |
Named Entity and Relation Extraction with Multi-Modal Retrieval (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing approaches to name entity recognition and relation extraction are knowledge-based and may not be highly relevant. |
| Approach: | They propose a multi-modal named entity recognition framework that leverages image information to improve the performance of NER and relation extraction. |
| Outcome: | The proposed framework can achieve state-of-the-art on four multi-modal named entity recognition datasets and one multi-module relation extraction dataset. |
Why Not Act on What You Know? Unleashing Safety Potential of LLMs via Self-Aware Guard Enhancement (2025.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated impressive capabilities across various tasks but are vulnerable to meticulously crafted jailbreak attacks. |
| Approach: | They propose a training-free defense strategy to align LLMs’ strong safety discrimination performance with their relatively weaker safety generation ability. |
| Outcome: | The proposed strategy achieves an average 99% success rate against numerous complex and covert jailbreak methods while maintaining helpfulness on general benchmarks. |
SWE-QA-Pro: A Representative Benchmark and Scalable Training Recipe for Repository-Level Code Understanding (2026.findings-acl)
Copied to clipboard
Songcheng Cai, Zhiheng Lyu, Yuansheng Ni, Xiangchao Chen, Baichuan Zhou, Shenzhe Zhu, Yi Lu, Haozhe Wang, Chi Ruan, Benjamin Schneider, Weixu Zhang, Xiang Li, Andy Zheng, Yuyu Zhang, Ping Nie, Wenhu Chen
| Challenge: | Existing benchmarks for agentic repository-level code understanding overlook long tail topics and rely on memorized knowledge. |
| Approach: | They propose a repository-level agentic code understanding benchmark that uses long-tail repositories with executable environments to enforce topical balance. |
| Outcome: | Empirically, a Qwen3-8B model trained with the proposed benchmark outperforms GPT-4o by 2.3 points. |
A Two-phase Prototypical Network Model for Incremental Few-shot Relation Classification (2020.coling-main)
Copied to clipboard
| Challenge: | Existing supervised and distantly supervised RC models ignore the emergence of novel relations in open environment. |
| Approach: | They propose a two-phase prototypical network with prototype attention alignment and triplet loss to dynamically recognize the novel relations with a few support instances without catastrophic forgetting. |
| Outcome: | Experiments show that the proposed model performs better on deep learning and few-shot learning . it can recognize the novel relations with a few support instances without catastrophic forgetting . |
Self-Reasoning Language Models: Unfold Hidden Reasoning Chains with Few Reasoning Catalyst (2025.findings-acl)
Copied to clipboard
| Challenge: | Recent studies have demonstrated that inference-time scaling increases performance of Large Language Models (LLMs) in various reasoning tasks such as mathematics and complex question answering by increasing the length of Chain-of-Thought (CoT). |
| Approach: | They propose a model which synthesizes longer CoT data and iteratively improves performance through self-training by incorporating a few demonstration examples. |
| Outcome: | The proposed model achieves an average improvement of more than +2.5 points across five reasoning tasks: MMLU, GSM8K, ARC-C, HellaSwag, and BBH on two backbone models. |
CoEvo: Coevolution of LLM and Retrieval Model for Domain-Specific Information Retrieval (2025.emnlp-main)
Copied to clipboard
Ang Li, Yiquan Wu, Yinghao Hu, Lizhi Qing, Shihang Wang, Chengyuan Liu, Tao Wu, Adam Jatowt, Ming Cai, Fei Wu, Kun Kuang
| Challenge: | Recent methods to enhance queries by generating intermediary elements can degrade retrieval performance . combining LLMs and retrievers can be difficult, resulting in unreliable or irrelevant intermediaries . |
| Approach: | They propose a framework that facilitates the coevolution of large language models and retrieval models. |
| Outcome: | The proposed framework facilitates the coevolution of LLMs and retrieval models. |
AgentGC: Evolutionary Learning-based Lossless Compression for Genomics Data with LLM-driven Multiple Agent (2026.findings-acl)
Copied to clipboard
Sun Hui, Ding Yanfeng, Huidong Ma, Chang Xu, Keyan Jin, Lizheng Zu, Cheng Zhong, Xiaoguang Liu, Gang Wang, Wentong Cai
| Challenge: | Lossless compression has made significant advancements in Genomics Data storage, sharing and management. |
| Approach: | They propose a novel agent-based GD Compressor with 3 layers with a multi-agent named Leader and Worker. |
| Outcome: | The proposed method improves on existing methods with low-level modeling and limited adaptability and user-unfriendly interface. |
AMoPO: Adaptive Multi-objective Preference Optimization without Reward Models and Reference Models (2025.findings-acl)
Copied to clipboard
Qi Liu, Jingqing Ruan, Hao Li, Haodong Zhao, Desheng Wang, Jiansong Chen, Wan Guanglu, Xunliang Cai, Zhi Zheng, Tong Xu
| Challenge: | Existing multi-objective preference alignment methods for large language models face limitations such as auxiliary reward/reference models and computational complexity. |
| Approach: | They propose a framework that achieves dynamic balance across preference dimensions by using dimension-aware generation metrics as implicit rewards. |
| Outcome: | Empirical results show that AMoPO outperforms state-of-the-art methods by 28.5% . |
Towards Verifiable Text Generation with Evolving Memory and Self-Reflection (2024.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) often produce factually incorrect information, also known as hallucination. |
| Approach: | They propose a framework for verifiable text generation with evolving memory and self-reflection that incorporates long-term memory to retain documents and recent documents. |
| Outcome: | The proposed framework outperforms baselines on five datasets across three knowledge-intensive tasks. |
MTR-Suite: A Framework for Evaluating and Synthesizing Conversational Retrieval Benchmarks (2026.acl-long)
Copied to clipboard
Junhao Ruan, Abudukeyumu Abudula, Bei Li, Yongjing Yin, Xinyu Liu, Kechen Jiao, Xin Chen, Jingang Wang, Xunliang Cai, Tong Xiao, JingBo Zhu
| Challenge: | Existing conversational retrieval benchmarks suffer from costly, sparse human annotation or rigid, unnatural automated heuristics. |
| Approach: | They propose a framework for auditing, synthesizing, and benchmarking conversational retrieval. |
| Outcome: | The proposed framework is based on three LLM-based auditors and a multi-agent system . it mimics production-style challenges (hard topic switching, verbosity) and offers superior discriminative power. |
Dynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Diffusion large language models (dLLMs) offer bidirectional attention and parallel generation . fixed anchors can enforce constraints, but they often impose rigid spans, leading to truncated reasoning . |
| Approach: | They propose a method that dynamically estimates end-anchor positions to adjust generation length before iterative infilling. |
| Outcome: | The proposed method improves format compliance and answer accuracy on GSM8K and MATH. |
KBGAN: Adversarial Learning for Knowledge Graph Embeddings (N18-1)
Copied to clipboard
| Challenge: | Existing knowledge graph embedding techniques lack the capability to access similarities between entities and relations. |
| Approach: | They propose an adversarial learning framework to improve knowledge graph embedding models . they use one knowledge graph embedded model as a negative sample generator . |
| Outcome: | The proposed framework improves the performance of knowledge graph embedding models on a link prediction task. |
A Corpus and Method for Chinese Named Entity Recognition in Manufacturing (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing resources and techniques for named entity recognition (NER) for manufacturing-specific named entities are limited. |
| Approach: | They propose a corpus of Chinese manufacturing specifications, named MS-NERC, with 4,424 sentences and 16,383 entities. |
| Outcome: | The proposed model outperforms neural methods in few-shot and rich-resource domains. |
EasyDistill: A Comprehensive Toolkit for Effective Knowledge Distillation of Large Language Models (2025.emnlp-demos)
Copied to clipboard
| Challenge: | In this paper, we present EasyDistill, a comprehensive toolkit designed for effective black-box and white-box knowledge distillation (KD) of large language models. |
| Approach: | They propose a toolkit for effective black-box and white-box knowledge distillation (KD) of large language models (LLMs). |
| Outcome: | The framework offers data synthesis, supervised fine-tuning, ranking optimization, and reinforcement learning techniques specifically tailored for KD scenarios. |
Large Language Models Meet Open-World Intent Discovery and Recognition: An Evaluation of ChatGPT (2023.emnlp-main)
Copied to clipboard
Xiaoshuai Song, Keqing He, Pei Wang, Guanting Dong, Yutao Mou, Jingang Wang, Yunsen Xian, Xunliang Cai, Weiran Xu
| Challenge: | Existing methods to fine-tune discriminative models address these challenges by focusing on in-domain intents. |
| Approach: | They evaluate ChatGPT on OOD intent discovery and generalized intent discovery tasks . they outline the strengths and weaknesses of ChatGPt and outline their results . |
| Outcome: | The proposed task aims to extend a closed intent classifier to open-world intent sets. |
Rethinking-based Code Summarization with Chain of Comments (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods focus on learning a direct mapping from pure code to summaries, overlooking the heterogeneity gap between code and summary. |
| Approach: | They propose a framework that uses chain of comments as auxiliary intermediate information to bridge the gap between code and summaries. |
| Outcome: | The proposed framework outperforms baseline models and multiple code Large Language Models by a large margin. |
UCL-Bench: A Chinese User-Centric Legal Benchmark for Large Language Models (2025.findings-naacl)
Copied to clipboard
Ruoli Gan, Duanyu Feng, Chen Zhang, Zhihang Lin, Haochen Jia, Hao Wang, Zhenyang Cai, Lei Cui, Qianqian Xie, Jimin Huang, Benyou Wang
| Challenge: | Existing legal benchmarks focusing on knowledge and logic evaluate LLMs on various tasks in legal domain, but few have explored the practical application of LLM by actual users. |
| Approach: | They propose a Chinese user-centric legal benchmark that aims to assess the practical application of LLMs by real users. |
| Outcome: | The proposed model outperforms existing models on various tasks in legal domain but does not outperfect ChatGPT. |
LLMs Know What They Need: Leveraging a Missing Information Guided Framework to Empower Retrieval-Augmented Generation (2025.coling-main)
Copied to clipboard
| Challenge: | Existing solutions to improve the accuracy of RAG are based on retrieval-augmented generation . however, RAG still faces several challenges in tackling complex multi-hop queries . |
| Approach: | They propose a Missing Information Guided Retrieve-Extraction-Solving paradigm that leverages the identification of missing information to generate a targeted query. |
| Outcome: | The proposed method can extract information from retrieved knowledge and know what is still missing. |
Scene Graph Enhanced Pseudo-Labeling for Referring Expression Comprehension (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Referring expression comprehension is a visual-linguistic task that involves localizing objects in images based on textual referring expressions. |
| Approach: | They propose a scene graph-based framework that generates high-quality pseudo region-query pairs . their method captures relationships between objects in images and generates expressions enriched with relation information. |
| Outcome: | The proposed framework outperforms existing methods by 10%, 12%, and 11% on RefCOCO, RefCoCO+, and Ref COCOg datasets. |
RICO: Improving Accuracy and Completeness in Image Recaptioning via Visual Reconstruction (2025.emnlp-main)
Copied to clipboard
Yuchi Wang, Yishuo Cai, Shuhuai Ren, Sihan Yang, Linli Yao, Yuanxin Liu, Yuanxing Zhang, Pengfei Wan, Xu Sun
| Challenge: | Existing recaptioning methods suffer from inaccuracies due to missing fine-grained details. |
| Approach: | They propose a framework that refines captions through visual reconstruction using a text-to-image model and a visual reconstruction framework. |
| Outcome: | The proposed framework outperforms baselines on CapsBench and CompreCap by 10%. |
Knowledge-Guided Cross-Topic Visual Question Generation (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing methods for visual question generation use answers or question types as constraints to generate questions. |
| Approach: | They propose a knowledge-guided cross-topic visual question generation task to generate unseen topics in cross-section scenarios. |
| Outcome: | The proposed model outperforms baselines and can generate unseen topic-related questions in cross-topic scenarios. |
TextBind: Multi-turn Interleaved Multimodal Instruction-following in the Wild (2024.findings-acl)
Copied to clipboard
| Challenge: | Large language models with instruction-following capabilities have revolutionized the field of artificial intelligence. |
| Approach: | They propose an annotation-free framework for empowering large language models with instruction-following capabilities. |
| Outcome: | The proposed framework generates multi-turn multimodal instruction-response conversations from a language model. |
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. |
Complex Question Decomposition for Semantic Parsing (P19-1)
Copied to clipboard
| Challenge: | Existing methods that ignore the decompositionality of complex questions are not suitable for complex question semantic parsing. |
| Approach: | They propose a hierarchical semantic parsing method which utilizes the decompositionality of complex questions for semantic paring. |
| Outcome: | The proposed method improves on a large scale complex question semantic parsing dataset. |
BAPO: Boundary-Aware Policy Optimization for Reliable Agentic Search (2026.findings-acl)
Copied to clipboard
Shiyu Liu, Yongjing Yin, Jianhao Yan, Yunbo Tang, Qinggang Zhang, Bei Li, Xin Chen, Jingang Wang, Xunliang Cai, Jinsong Su
| Challenge: | Existing RL-based agentic search models fail to recognize reasoning boundaries and rarely admit "I DON'T KNOW" lack of reliability leads to plausible but unreliable answers, introducing significant risks . |
| Approach: | They propose a framework to cultivate reliable boundary awareness without compromising accuracy. |
| Outcome: | Experiments show that the proposed framework improves the reliability of agentic search models. |
Too Consistent to Detect: A Study of Self-Consistent Errors in LLMs (2025.emnlp-main)
Copied to clipboard
Hexiang Tan, Fei Sun, Sha Liu, Du Su, Qi Cao, Xin Chen, Jingang Wang, Xunliang Cai, Yuanzhuo Wang, Huawei Shen, Xueqi Cheng
| Challenge: | Existing detection methods fail to account for **self-consistent error** . study identifies self-consistency errors and evaluates them . |
| Approach: | They propose a method that fuses hidden state evidence from an external verifier LLM to detect self-consistent errors. |
| Outcome: | The proposed method significantly enhances performance on self-consistent errors across three LLM families. |
The World is Not Binary: Learning to Rank with Grayscale Data for Dialogue Response Selection (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing approaches to learning-to-rank response selection are suboptimal due to ignorance of diversity of response quality. |
| Approach: | They propose to use off-the-shelf response retrieval models as automatic grayscale data generators to train response selection models. |
| Outcome: | The proposed approach can be automated without human effort on grayscale data. |
EULoInf: Efficient Hessian-Free Entropy Based Uncertainty-Aware Data Influence Approximation (2026.findings-acl)
Copied to clipboard
| Challenge: | Extensive studies show that the effectiveness of fine-tuning heavily relies on the quality of training data. |
| Approach: | They propose a framework that approximates influence via uncertainty and gradient based validation loss lookahead. |
| Outcome: | The proposed framework matches or outperforms prior methods across diverse tasks and LLM architectures while reducing computational time and memory usage by over 50%. |
A Decoupled Multi-Agent Framework for Complex Text Style Transfer (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing models for text style transfer struggle with complex styles . existing models perform well on simple styles like sentiment and formality . |
| Approach: | They propose a multi-agent self-check framework that includes a large language model as a planner for disentangling subtasks and expert agents for executing the subtask. |
| Outcome: | The proposed framework significantly improves style strength and content preservation on simple and complex style datasets. |
FIRE: Flexible Integration of Data Quality Ratings for Effective Pretraining (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods to evaluate data quality rely on heuristic techniques or single quality signals. |
| Approach: | They propose a framework for integrating multiple data quality raters that integrates multiple quality signals into a unified space and provides a comprehensive quality signal for each data point. |
| Outcome: | The proposed framework outperforms existing methods and boosts model performance across a wide range of downstream tasks while requiring less than 37.5% tokens to reach the target performance. |
AJ-Bench: Benchmarking Agent-as-a-Judge for Environment-Aware Evaluation (2026.findings-acl)
Copied to clipboard
Wentao Shi, Yu Wang, Yuyang Zhao, Yuxin Chen, Fuli Feng, Xueyuan Hao, Xi Su, Qi GU, Hui Su, Xunliang Cai, Xiangnan He
| Challenge: | Existing approaches to verify agent behaviors in complex environments rely on rule-based verifiers or LLM-as-a-Judge models. |
| Approach: | They propose a benchmark to evaluate Agent-as-a-Judge across three domains . the benchmark covers search, data systems, and graphical user interfaces - with 155 tasks and 516 trajectories . |
| Outcome: | The proposed benchmark outperforms existing benchmarks in search, data systems, and GUI domains while revealing open challenges in agent-based verification. |
Cross-lingual Multimodal Sentiment Analysis for Low-Resource Languages via Language Family Disentanglement and Rethinking Transfer (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing multimodal sentiment analysis methods are limited to textual data and cannot handle multimodal scenarios. |
| Approach: | They propose a transfer learning framework that allows cross-lingual and cross-modal alignments and a language family disentanglement module that enhances the sharing of language universals within families. |
| Outcome: | The proposed method is superior to existing methods and can handle low-resource languages. |
Observations and Remedies for Large Language Model Bias in Self-Consuming Performative Loop (2026.acl-long)
Copied to clipboard
| Challenge: | Existing synthetic training loops for large language models cause performance drops and induce emerging biases . a large amount of generated content is posted to coding platforms, social media platforms and other platforms on the internet . |
| Approach: | They propose a self-consuming retraining loop where models are trained on their own outputs . they use a control loop to isolate and analyze feedback-driven bias evolution . |
| Outcome: | The proposed model increases preference bias and decreases disparate bias. |
Improve Decoding Factuality by Token-wise Cross Layer Entropy of Large Language Models (2025.findings-naacl)
Copied to clipboard
| Challenge: | Large language models (LLMs) often struggle with the issue of generating inaccurate or fabricated content even when they possess correct knowledge. |
| Approach: | They propose a decoding method that mitigates hallucinations without extra training . they propose entropy eNhanced decoding that leverages inner probability changes . |
| Outcome: | The proposed method improves the truthfulness and informativeness of generation while maintaining robust QA accuracy. |
Beyond the Known: Investigating LLMs Performance on Out-of-Domain Intent Detection (2024.lrec-main)
Copied to clipboard
Pei Wang, Keqing He, Yejie Wang, Xiaoshuai Song, Yutao Mou, Jingang Wang, Yunsen Xian, Xunliang Cai, Weiran Xu
| Challenge: | Out-of-domain (OOD) intent detection is crucial for task-oriented dialogue systems. |
| Approach: | They conduct a comprehensive evaluation of large language models (LLMs) under various experimental settings and outline their strengths and weaknesses. |
| Outcome: | The proposed models exhibit strong zero-shot and few-shot capabilities, but is still at a disadvantage compared to models fine-tuned with full resource. |
DataSciBench: An LLM Agent Benchmark for Data Science (2026.findings-acl)
Copied to clipboard
Dan Zhang, Sining Zhoubian, Min Cai, Fengzu Li, Lekang Yang, Wei Wang, Tianjiao Dong, Ziniu Hu, Jie Tang, Yisong Yue
| Challenge: | Existing benchmarks focus on single task, simple evaluation metrics, and readily available ground truth (GT) DataSciBench is built on curated, natural, and challenging prompts with complex evaluation criteria and uncertain GT. |
| Approach: | They propose a benchmark for evaluating Large Language Models in data science that integrates LLM-based self-consistency and human verification to ensure accuracy. |
| Outcome: | The proposed framework outperforms open-source models in all metrics and offers rigorous insights into LLM strengths and weaknesses. |
Skeleton-to-Response: Dialogue Generation Guided by Retrieval Memory (N19-1)
Copied to clipboard
| Challenge: | Existing generative dialogue models generate responses from input queries . however, the results are limited and the models are unsatisfactory . |
| Approach: | They propose a framework which exploits retrieval results via a skeleton-to-response paradigm . they extract a query skelet and use it to generate a new skele and response . |
| Outcome: | The proposed approach significantly improves the informativeness of the generated responses. |
MatchTIR: Fine-Grained Supervision for Tool-Integrated Reasoning via Bipartite Matching (2026.acl-long)
Copied to clipboard
| Challenge: | Existing reinforcement learning methods rely on outcome- or trajectory-level rewards, assigning uniform advantages to all steps within a trajectory. |
| Approach: | They propose a framework that introduces fine-grained supervision via bipartite matching-based turn-level reward assignment and dual-level advantage estimation. |
| Outcome: | The proposed framework surpasses the majority of 8B competitors on three benchmarks. |
Tricking Retrievers with Influential Tokens: An Efficient Black-Box Corpus Poisoning Attack (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing methods for crafting adversarial passages are slow and computationally expensive, requiring either access to retriever’s gradients or large computational resources. |
| Approach: | They propose a method that leverages two key characteristics of retrievers: insensitivity to token order and bias towards influential tokens to generate effective adversarial passages. |
| Outcome: | The proposed method achieves superior efficiency and scalability compared to existing methods while maintaining comparable or better attack success rates across multiple datasets. |
SegTune: Structured and Fine-Grained Control for Song Generation (2026.acl-long)
Copied to clipboard
Yuejiao Wang, Zihao Ji, Pengfei Cai, Xu Li, Haorui Zheng, Zewen Song, Zhongliang Liu, Chen Zhang, Pengfei Wan
| Challenge: | Recent advances in neural song generation have enabled high-quality synthesis from lyrics and global textual prompts. |
| Approach: | They propose a framework that allows users to specify local musical descriptions aligned to song segments. |
| Outcome: | The proposed framework outperforms baselines in musicality and controllability. |
LLM-FK: Multi-Agent LLM Reasoning for Foreign Key Detection in Large-Scale Complex Databases (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for detecting missing foreign keys are limited in capturing semantic dependencies across schemas. |
| Approach: | They propose a framework that integrates four agents to detect missing foreign keys . they propose combinatorial search space explosion, ambiguous inference and global inconsistency . |
| Outcome: | The proposed framework achieves F1-scores above 93% on large-scale MusicBrainz database . it reduces candidate search space by two to three orders of magnitude without losing true FKs . |
MDC-Bench: A Multidisciplinary Causal Benchmark Based on Causal Structures for Evaluating Large Language Models (2026.findings-acl)
Copied to clipboard
Peng Wang, Yuxiong Yan, Xiao Ding, Kai Xiong, Bibo Cai, Chao Peng, Yutai Hou, Dandan Tu, Bing Qin, Ting Liu
| Challenge: | Existing causal datasets focus on the commonsense domain, but LLMs perform poorly when answering complex questions. |
| Approach: | They propose a multidisciplinary causal evaluation benchmark to assess LLMs' knowledge and skills. |
| Outcome: | The proposed model improves in domain specialization, structural diversity, and task complexity. |
Retrieval-based Knowledge Transfer: An Effective Approach for Extreme Large Language Model Compression (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Large-scale pre-trained language models (LLMs) have demonstrated exceptional performance in various natural language processing (NLP) tasks. |
| Approach: | They propose a new compression paradigm that extracts knowledge from pre-trained language models to construct a knowledge store from which the model can leverage it for effective inference. |
| Outcome: | The proposed model extracts knowledge from LLMs to construct a knowledge store, which the model can leverage for effective inference. |
WebCPM: Interactive Web Search for Chinese Long-form Question Answering (2023.acl-long)
Copied to clipboard
Yujia Qin, Zihan Cai, Dian Jin, Lan Yan, Shihao Liang, Kunlun Zhu, Yankai Lin, Xu Han, Ning Ding, Huadong Wang, Ruobing Xie, Fanchao Qi, Zhiyuan Liu, Maosong Sun, Jie Zhou
| Challenge: | Long-form question answering requires two procedures: information retrieval and information synthesis. |
| Approach: | They propose a Chinese long-form question answering dataset called WebCPM . the dataset is based on a web search interface that engages with a search engine in real time . |
| Outcome: | The proposed dataset generates answers that are no worse than human-written ones . the dataset is the first Chinese LFQA dataset . |
Uplift-RAG: Uplift-Driven Knowledge Preference Alignment for Retrieval-Augmented Generation (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing efforts to estimate document utility rely on downstream generation performance, which conflates the influence of external documents with the intrinsic knowledge of the LLM. |
| Approach: | They propose an uplift-based definition of document utility that quantifies each document’s marginal benefit over the LLM’s internal knowledge. |
| Outcome: | The proposed framework improves the performance of the LLM by incorporating external retrieved documents into the model. |
Neural Machine Translation with Monolingual Translation Memory (2021.acl-long)
Copied to clipboard
| Challenge: | Existing work has shown that Translation Memory (TM) can boost the performance of Neural Machine Translation (NMT) |
| Approach: | They propose a framework that uses monolingual memory and performs learnable memory retrieval in a cross-lingual manner. |
| Outcome: | The proposed framework outperforms strong TM-augmented NMT baselines using bilingual TM and outperformed existing models in low-resource and domain adaptation scenarios. |
The Agent’s First Day: Benchmarking Learning, Exploration, and Scheduling in the Workplace Scenarios (2026.findings-acl)
Copied to clipboard
Daocheng Fu, Jianbiao Mei, Rong Wu, Xuemeng Yang, Jia Xu, Ding Wang, Pinlong Cai, Yong Liu, Licheng Wen, Botian Shi
| Challenge: | Existing research mainly focuses on performance upper bounds in static environments, overlooking stochastic real-world deployment. |
| Approach: | They propose a dynamic evaluation environment that simulates a "trainee" agent continuously exploring a novel setting. |
| Outcome: | The proposed model evaluates agents in a dynamic evaluation environment that simulates a "trainee" agent continuously exploring a novel setting. |
FanChuan: A Multilingual and Graph-Structured Benchmark For Parody Detection and Analysis (2025.findings-acl)
Copied to clipboard
Yilun Zheng, Sha Li, Fangkun Wu, Yang Ziyi, Lin Hongchao, Zhichao Hu, Cai Xinjun, Ziming Wang, Jinxuan Chen, Sitao Luan, Jiahao Xu, Lihui Chen
| Challenge: | Parody is an emerging phenomenon on social media, where individuals imitate a role or position opposite to their own . limited available data and deficient diversity in current datasets hinder study of parody . |
| Approach: | They build a dataset of parody users and annotated comments from both English and Chinese corpora to test parody detection and comment sentiment analysis. |
| Outcome: | The proposed datasets provide richer contextual information, which is lacking in existing datasets. |
Can Multimodal LLMs See Materials Clearly? A Multimodal Benchmark on Materials Characterization (2025.findings-emnlp)
Copied to clipboard
Zhengzhao Lai, Youbin Zheng, Zhenyang Cai, Haonan Lyu, Jingpu Yang, Hong-Qing Liang, Yan Hu, Benyou Wang
| Challenge: | characterization imaging data is fundamental to acquiring materials information. |
| Approach: | a team of researchers develop a benchmark for materials characterization imaging data . the goal is to bridge this gap by addressing 1,500 questions that require expert-level expertise. |
| Outcome: | a new benchmark for materials characterization imaging data is presented . the benchmark reveals that MLLMs perform poorly when addressing higher-level questions . |
Con-ReCall: Detecting Pre-training Data in LLMs via Contrastive Decoding (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods analyze training data with member and non-member contexts, overlooking potential insights from both member and not-member. |
| Approach: | They propose a method that leverages asymmetric distributional shifts induced by member and non-member contexts through contrastive decoding to enhance membership inference. |
| Outcome: | The proposed approach outperforms the current state-of-the-art on the WikiMIA benchmark and is robust against various text manipulation techniques. |
Entity-centered Cross-document Relation Extraction (2022.emnlp-main)
Copied to clipboard
Fengqi Wang, Fei Li, Hao Fei, Jingye Li, Shengqiong Wu, Fangfang Su, Wenxuan Shi, Donghong Ji, Bo Cai
| Challenge: | Existing methods for relation extraction only use text snippets surrounding target entities in multiple documents. |
| Approach: | They propose a relation-extraction model that uses cross-path entity relation attention to detect the semantic relations between entities in a given text. |
| Outcome: | The proposed method outperforms the state-of-the-art methods in the dataset CodRED by 10%. |
ALYMPICS: LLM Agents Meet Game Theory (2025.coling-main)
Copied to clipboard
| Challenge: | Alympics provides a framework for simulating human-like strategic interactions with Large Language Model (LLM) agents. |
| Approach: | They propose a framework utilizing Large Language Models (LLM) agents for empirical game theory research. |
| Outcome: | The proposed framework can be used to study human-like strategic interactions with large language model (LLM) agents in a game on the multi-round auction of scarce survival resources. |
Can LLMs Act as Historians? Evaluating Historical Research Capabilities of LLMs via the Chinese Imperial Examination (2026.acl-long)
Copied to clipboard
Lirong Gao, Zeqing Wang, Yuyan Cai, Jiayi Deng, Yanmei Gu, Yiming Zhang, Jia Zhou, Yanfei Zhang, Junbo Zhao
| Challenge: | Existing benchmarks assess basic knowledge breadth or lexical understanding, failing to capture higher-order skills that are central to historical research. |
| Approach: | They propose a benchmark anchored in the Chinese Imperial Examination system that assesses historical knowledge and lexical understanding. |
| Outcome: | The new benchmark aims to assess the ability of LLMs to process historical materials and documents. |
COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human Values (2026.findings-eacl)
Copied to clipboard
Siwei Wu, JinCheng Ren, Xeron Du, Shuyue Guo, Xingwei Qu, Yiming Liang, Jie Liu, Yunwen Li, Tyler Loakman, Tianyu Zheng, Boyu Feng, Huaqing Yuan, Zili Wang, Jiaheng Liu, Wenhao Huang, Chenglin Cai, Haoran Que, Jian Yang, Yuelin Bai, Zekun Moore Wang, Zhouliang Yu, Qunshu Lin, Ding Pan, Yuchen Eleanor Jiang, Tiannan Wang, Wangchunshu Zhou, Shenzhi Wang, Xingyuan Bu, Minghao Liu, Guoyin Wang, Ge Zhang, Chenghua Lin
| Challenge: | Existing Chinese preference datasets suffer from limited scale, restricted domain coverage, and insufficiently rigorous data validation. |
| Approach: | They propose an LLM-based data annotation pipeline with no human intervention to annotate Chinese preference datasets. |
| Outcome: | The proposed pipeline outperforms existing Chinese preference datasets on AlignBench and Chinese Reward Benchmark. |
CTR-Guided Generative Query Suggestion in Conversational Search (2025.emnlp-industry)
Copied to clipboard
Erxue Min, Hsiu-Yuan Huang, Xihong Yang, Min Yang, Xin Jia, Yunfang Wu, Hengyi Cai, Junfeng Wang, Shuaiqiang Wang, Dawei Yin
| Challenge: | Generating effective query suggestions requires aligning model outputs with user click preferences. |
| Approach: | They propose a generative framework that leverages click modeling to denoise implicit feedback and enables reliable preference optimization for improving real-world user engagement. |
| Outcome: | The proposed framework outperforms strong baselines in CTR, relevance, diversity and diversity. |
SMedBERT: A Knowledge-Enhanced Pre-trained Language Model with Structured Semantics for Medical Text Mining (2021.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge-based PLMs are based on linked-entity information, but they only use linked-enemy information as auxiliary information. |
| Approach: | They propose to integrate semantic knowledge from neighbours of linked-entity into a medical PLM that integrates heterogeneous-entities into the homogeneously neighbouring entity structure. |
| Outcome: | Experiments show that SMedBERT outperforms baselines in knowledge-intensive Chinese medical tasks. |
SeDev: Structured Semantic Exploration for LLM-Driven Code Generation (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown remarkable capabilities in automating code generation, but they suffer from insufficient exploration of the vast solution space. |
| Approach: | They propose a large-scale LLM-driven code generation framework that efficiently finds high-quality solutions in only a few iterations. |
| Outcome: | The proposed framework outperforms baselines while maintaining reasonable time and computational costs. |
Chat-Driven Text Generation and Interaction for Person Retrieval (2025.emnlp-main)
Copied to clipboard
| Challenge: | Text-based person search (TBPS) aims to retrieve images of a target individual from large-scale databases using natural language descriptions. |
| Approach: | They propose two modules to improve retrieval accuracy and usability . they generate rich pseudo-labels through simulated dialogues with MLLMs . |
| Outcome: | The proposed method achieves competitive or superior results while eliminating the need for manual captions. |
Effidit: An Assistant for Improving Writing Efficiency (2023.acl-demo)
Copied to clipboard
Shuming Shi, Enbo Zhao, Wei Bi, Deng Cai, Leyang Cui, Xinting Huang, Haiyun Jiang, Duyu Tang, Kaiqiang Song, Longyue Wang, Chenyan Huang, Guoping Huang, Yan Wang, Piji Li
| Challenge: | Effidit is a digital writing assistant that provides three modules to help users write faster and more efficiently. |
| Approach: | They present Effidit, a digital writing assistant that provides three modules to help users write higher-quality text more efficiently. |
| Outcome: | Effidit expands the capabilities of a typical writing assistant by providing three modules . Effit can help users create their own text faster and more efficiently . |
APP: Adaptive Prototypical Pseudo-Labeling for Few-shot OOD Detection (2023.findings-emnlp)
Copied to clipboard
Pei Wang, Keqing He, Yutao Mou, Xiaoshuai Song, Yanan Wu, Jingang Wang, Yunsen Xian, Xunliang Cai, Weiran Xu
| Challenge: | Existing methods for detecting out-of-domain (OOD) intents are hard to label . previous studies use labeled in-domain data to learn intent representations . |
| Approach: | They propose a prototypical pseudo-labeling method for few-shot OOD detection . they propose 'protoOOD' framework and adaptive pseudo-labeled method . |
| Outcome: | The proposed method is able to detect out-of-domain (OOD) intents from user queries. |
League of LLMs: A Benchmark-Free Paradigm for Mutual Evaluation of Large Language Models (2026.acl-long)
Copied to clipboard
Qianhong Guo, Wei Xie, Xiaofang Cai, Enze Wang, Shuoyoucheng Ma, Xiaobing Sun, Tian Xia, Kai Chen, Xiaofeng Wang, Baosheng Wang
| Challenge: | Large language models (LLMs) have shown exceptional capabilities across a wide range of tasks, but reliable evaluation remains a challenge due to data contamination, opaque operation, and subjective preferences. |
| Approach: | They propose a benchmark-free evaluation paradigm that organizes multiple LLMs into a self-governed league for multi-round mutual evaluation. |
| Outcome: | Experiments on eight mainstream LLMs in mathematics and programming show that the proposed model can distinguish capabilities while maintaining high internal ranking stability. |
ICG: Improving Cover Image Generation via MLLM-based Prompting and Personalized Preference Alignment (2025.emnlp-main)
Copied to clipboard
Zhipeng Bian, Jieming Zhu, Qijiong Liu, Wang Lin, Guohao Cai, Zhaocheng Du, Jiacheng Sun, Zhou Zhao, Zhenhua Dong
| Challenge: | Large language models and diffusion models have opened new possibilities for AI-generated content . personalized cover image generation remains underexplored despite its critical role in boosting user engagement on digital platforms. |
| Approach: | They propose a framework that integrates MLLM-based prompting with personalized preference alignment to generate high-quality, contextually relevant covers. |
| Outcome: | The proposed framework improves image quality, semantic fidelity, and personalization, leading to stronger user appeal and offline recommendation accuracy in downstream tasks. |
Prejudge-Before-Think: Enhancing Large Language Models at Test-Time by Process Prejudge Reasoning (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Experimental results from competition-level complex reasoning demonstrate that bootstrapping with process prejudge can significantly enhance the reasoning ability of LLMs. |
| Approach: | They propose a new process prejudge strategy for LLM reasoning that bootstraps with process prejudgment . |
| Outcome: | The proposed method can be bootstrapped with process prejudge in LLM reasoning . it allows the model to anticipate errors rather than relying on trial and error. |
Multi-Scale Distribution Deep Variational Autoencoder for Explanation Generation (2022.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for generating explanations for recommender systems produce generic explanations that fail to incorporate user and item specific details. |
| Approach: | They propose a multi-scale distribution deepvariational autoencoder with a prior network that eliminates noise while retaining meaningful signals in the input. |
| Outcome: | The proposed models can generate explanations with concrete input-specific contents. |
MobileWorld: Benchmarking Autonomous Mobile Agents in Agent-User Interactive and MCP-Augmented Environments (2026.acl-long)
Copied to clipboard
Quyu Kong, Xu Zhang, Zhenyu Yang, Nolan Gao, Chen Liu, Panrong Tong, Chenglin Cai, Hanzhang Zhou, Jianan Zhang, Liangyu Chen, Zhidan Liu, Steven Hoi, Yue Wang
| Challenge: | AndroidWorld is the dominant mobile GUI agent evaluation benchmark, but its success rates are low . despite reproducible emulator environment, it lacks key application categories such as e-commerce and enterprise communication. |
| Approach: | They propose a benchmark for mobile GUI agents that reflects real-world usage through long-horizon, cross-application workflows. |
| Outcome: | The proposed framework achieves over 90% success rates, while AndroidWorld is the dominant benchmark. |
Rethinking the Reversal Curse of LLMs: a Prescription from Human Knowledge Reversal (2024.emnlp-main)
Copied to clipboard
Zhicong Lu, Li Jin, Peiguang Li, Yu Tian, Linhao Zhang, Sirui Wang, Guangluan Xu, Changyuan Tian, Xunliang Cai
| Challenge: | Existing methods for large language models (LLMs) are limited by their aggressive sample permutation and lack a detailed understanding of the underlying reasons for the reversal curse. |
| Approach: | They propose a method which enhances bidirectional entity correlation modeling and pairwise relationship reasoning to overcome the reversal curse. |
| Outcome: | The proposed method overcomes the reversal curse by augmenting the samples with entity order-reversals and semantically preserved question-answer pairs. |
PCA-Bench: Evaluating Multimodal Large Language Models in Perception-Cognition-Action Chain (2024.findings-acl)
Copied to clipboard
Liang Chen, Yichi Zhang, Shuhuai Ren, Haozhe Zhao, Zefan Cai, Yuchi Wang, Peiyi Wang, Xiangdi Meng, Tianyu Liu, Baobao Chang
| Challenge: | a new multimodal decision-making benchmark evaluates the integrated capabilities of multimodal large language models. |
| Approach: | They propose a multimodal decision-making benchmark for evaluating MLLMs . they propose an automatic evaluation protocol to assess 10 prevalent ML models . |
| Outcome: | The proposed benchmark improves performance of multimodal large language models in three scenarios . the model is required to integrate multiple capabilities to make accurate decisions . |
VisFinEval: A Scenario-Driven Chinese Multimodal Benchmark for Holistic Financial Understanding (2025.emnlp-main)
Copied to clipboard
Zhaowei Liu, Xin Guo, Haotian Xia, Lingfeng Zeng, Fangqi Lou, Jinyi Niu, Mengping Li, Qi Qi, Jiahuan Li, Wei Zhang, Yinglong Wang, Weige Cai, Weining Shen, Liwen Zhang
| Challenge: | Existing benchmarks focus on text comprehension, but MLLMs lack the ability to integrate visual data over financial visuals. |
| Approach: | They evaluate 21 state-of-the-art multimodal large language models in a zero-shot setting . they use an annotated question–answer pair from eight common financial image modalities . |
| Outcome: | The new benchmark outperforms existing models but trailed financial experts by 14 percentage points. |
On the Representation Geometry of LoRA Model Merging (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for low-rank Adaptation (LoRA) fine-tuning focus on globally shared structure . combining SVD with CUR improves performance of LoRA model merging . |
| Approach: | They propose a training-free method that combines SVD and CUR decomposition to improve LoRA merging performance. |
| Outcome: | The proposed procedure improves on vision and language benchmarks. |
Fine-Grained Features-based Code Search for Precise Query-Code Matching (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods to locate code snippets from databases represent the semantics of code and query by averaging the features of each token and word. |
| Approach: | They propose a fine-grained code search model that consists of a cross-modal encoder, mapping layer and classification layer to capture fine-granular interactions between code and query. |
| Outcome: | The proposed model significantly outperforms existing methods across multiple programming language datasets. |
CCG: Rare-Label Prediction via Neural SEM–Driven Causal Game (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Multi-label classification (MLC) faces persistent challenges from label imbalance, spurious correlations, distribution shifts, especially in rare label prediction. |
| Approach: | They propose a Causal Cooperative Game framework that models multi-player cooperative process for multi-label classification. |
| Outcome: | The proposed framework improves rare label prediction and overall robustness compared to baselines. |
XDailyDialog: A Multilingual Parallel Dialogue Corpus (2023.acl-long)
Copied to clipboard
Zeming Liu, Ping Nie, Jie Cai, Haifeng Wang, Zheng-Yu Niu, Peng Zhang, Mrinmaya Sachan, Kaiping Peng
| Challenge: | Existing datasets for open-domain dialogue modeling limited to a single language . absence of multilingual datasets hinders development of robust open- domain dialog systems . |
| Approach: | They propose a multilingual parallel open-domain dialog dataset to explore multilingual and cross-lingual open- domain dialog. |
| Outcome: | The proposed model can be used to explore multilingual and cross-lingual open-domain dialogs in other languages. |
Dynamic Fisher-weighted Model Merging via Bayesian Optimization (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing merging approaches involve scaling the parameters model-wise or integrating parameter importance parameter-wise. |
| Approach: | They propose a method for merging model-based models at the parameter level without training data or joint training. |
| Outcome: | The proposed model merging framework outperforms baseline models on validation sets. |
Making Every Step Effective: Jailbreaking Large Vision-Language Models Through Hierarchical KV Equalization (2025.findings-emnlp)
Copied to clipboard
| Challenge: | HKVE selectively accepts gradient optimization results based on the distribution of attention scores across different layers, ensuring that every optimization step positively contributes to the attack. |
| Approach: | They propose a framework that selectively accepts gradient optimization results based on the distribution of attention scores across different layers and selectively takes them into account when calculating the attack success rate. |
| Outcome: | The proposed framework outperforms existing methods by achieving success rates of 75.08% on MiniGPT4, 85.84% on LLaVA and 81.00% on Qwen-VL. |
DiMo-GUI: Advancing Test-time Scaling in GUI Grounding via Modality-Aware Visual Reasoning (2025.emnlp-main)
Copied to clipboard
| Challenge: | DiMo-GUI is a training-free framework for GUI grounding that splits input into textual elements and iconic elements, allowing the model to reason over each modality independently using general-purpose vision-language models. |
| Approach: | They propose a training-free framework for GUI grounding that leverages two core strategies: dynamic visual grounding and modality-aware optimization. |
| Outcome: | The proposed framework splits the input into textual elements and iconic elements, allowing the model to reason over each modality independently using general-purpose vision-language models. |
Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text (2026.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) can be used to effectively utilize tools in multi-turn interactions, but acquiring diverse and realistic multi-step tool-use data remains a challenge. |
| Approach: | They propose a text-based data synthesis pipeline that generates multi-turn tool-use trajectories from text corpora using relevance filtering, workflow tool extraction, trajectory grounding, and complexity refinement. |
| Outcome: | The proposed model achieves 14.9% improvement on the BFCL V3 Multi-turn benchmark while significantly reducing inference latency and costs. |
Speculative Decoding via Early-exiting for Faster LLM Inference with Thompson Sampling Control Mechanism (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to generate draft tokens in large language models are expensive and resource-intensive. |
| Approach: | They propose an approach to generate draft tokens using a segment of the LLM and a self-distillation method to enhance the quality of draft token. |
| Outcome: | The proposed approach generates draft tokens using a segment of the LLM and a self-distillation method to improve quality and speed up generation. |
Reasoning with OmniThought: A Large CoT Dataset with Verbosity and Cognitive Difficulty Annotations (2026.acl-long)
Copied to clipboard
| Challenge: | Existing resources often fail to provide extensive reasoning problems with coherent CoT processes distilled from multiple teacher models. |
| Approach: | They propose a large-scale dataset featuring 2 million CoT processes generated by multiple powerful LRMs. |
| Outcome: | The proposed dataset features 2 million CoT processes and is validated by multiple powerful LRMs. |
Flexibly Utilize Memory for Long-Term Conversation via a Fragment-then-Compose Framework (2025.emnlp-main)
Copied to clipboard
Cai Ke, Yiming Du, Bin Liang, Yifan Xiang, Lin Gui, Zhongyang Li, Baojun Wang, Yue Yu, Hui Wang, Kam-Fai Wong, Ruifeng Xu
| Challenge: | Large language models extract useful information from conversation history to enhance the response in long-term conversations. |
| Approach: | They propose a Fragment-then-Compose framework to optimize memory utilization for long-term open-domain conversation. |
| Outcome: | The proposed framework can be used to extract useful information from conversation history . it can be adapted to different situations and improve response generation . |
Provably Safe Offline-to-Online RL: Decoupling Learning from Data-Driven Safety Enforcement (2026.acl-long)
Copied to clipboard
| Challenge: | Hybrid offline–online reinforcement learning (O2O RL) promises both sample efficiency and robust exploration, but suffers from instability due to distribution shift between offline and online data. |
| Approach: | They propose a framework that decouples policy optimization from safety enforcement . they propose dynamic curricula that gradually extend temporal horizons and anneal offline–online data mixing . |
| Outcome: | The proposed framework preserves the exploratory value of online interactions without collapsing to conservative policies. |
MavenCoder: Competitive Code Generation via Model Adaptive Planning Strategies and Multi-Perspective Verification Enhancement (2026.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have significantly enhanced automated program synthesis. |
| Approach: | They propose a model-adaptive and verification–enhanced framework for competition-level code generation that leverages adaptive assessment aligned with the model’s capabilities to select planning strategies while providing timely feedback and correction via multi-perspective verification. |
| Outcome: | The proposed framework outperforms existing state-of-the-art approaches on livecodebench, humanEval+, MBPP+, and codecontests, and achieves pass@1 results exceeding 3%–40%. |
U-Fold: Dynamic Intent-Aware Context Folding for User-Centric Agents (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing context-folding methods are designed for single-query or single-intent scenarios. |
| Approach: | They propose a dynamic context-folding framework tailored to user-centric tasks that preserves fine-grained information through dynamic context folding. |
| Outcome: | The proposed framework outperforms ReAct and previous folding frameworks on long, noisy tasks. |
Generalized Category Discovery with Large Language Models in the Loop (2024.findings-acl)
Copied to clipboard
Wenbin An, Wenkai Shi, Feng Tian, Haonan Lin, QianYing Wang, Yaqiang Wu, Mingxiang Cai, Luyan Wang, Yan Chen, Haiping Zhu, Ping Chen
| Challenge: | Generalized Category Discovery (GCD) is a crucial task that aims to recognize both known and novel categories from a set of unlabeled data. |
| Approach: | They propose a framework that introduces Large Language Models into the training loop to generate category names without human effort. |
| Outcome: | The proposed framework outperforms SOTA models on three benchmark datasets and generates accurate category names for the discovered clusters. |
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. |
FlightGPT: Towards Generalizable and Interpretable UAV Vision-and-Language Navigation with Vision-Language Models (2025.emnlp-main)
Copied to clipboard
Hengxing Cai, Jinhan Dong, Jingjun Tan, Jingcheng Deng, Sihang Li, Zhifeng Gao, Haidong Wang, Zicheng Su, Agachai Sumalee, Renxin Zhong
| Challenge: | Existing methods for vision-and-language navigation struggle with insufficient multimodal fusion, weak generalization, and poor interpretability. |
| Approach: | They propose a framework for UAV vision-and-language navigation that integrates natural language instructions with visual observations to improve multimodal fusion and interpretability. |
| Outcome: | The proposed framework achieves state-of-the-art performance across all scenarios, with a 9.22% higher success rate than the strongest baseline in unseen environments. |
Preference Curriculum: LLMs Should Always Be Pretrained on Their Preferred Data (2025.findings-acl)
Copied to clipboard
Xuemiao Zhang, Xu Liangyu, Feiyu Duan, Yongwei Zhou, Sirui Wang, Rongxiang Weng, Jingang Wang, Xunliang Cai
| Challenge: | Existing methods of uniformly sampling data throughout the pretraining process are suboptimal because they overlook the model's evolving data preferences. |
| Approach: | They propose a Perplexity Difference (PD) based Preference Curriculum learning framework which perceives and uses the data preferred by LLMs as their capabilities improve . they propose PDPC to complete the arrangement of the dataset offline and ensure continuous training without interruption. |
| Outcome: | The proposed framework surpasses baselines on 1.3B and 3B models and achieves an increased average accuracy of over 8.1% across MMLU and CMMLU. |
SyncThink: A Training-Free Strategy to Align Inference Termination with Reasoning Saturation (2026.findings-acl)
Copied to clipboard
| Challenge: | Large language models (LLMs) achieve strong reasoning with Chain-of-Thought prompting, but long and redundant traces substantially increase inference cost. |
| Approach: | They propose a training-free and plug-and-play decoding method that reduces CoT overhead without modifying model weights. |
| Outcome: | Experiments on GSM8K, MMLU, GPQA, and BBH show that SyncThink achieves 62.00% average Top@1 accuracy using 656 generated tokens and 28.68s latency, compared to 61.22%, 2141 tokens, and 92.01s for full CoT decoding. |
In-context Learning for Few-shot Multimodal Named Entity Recognition (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for named entity recognition are time-consuming and laborintensive. |
| Approach: | They propose a few-shot multimodal named entity recognition task that uses few examples to locate and identify named entities for a text-image pair. |
| Outcome: | The proposed framework outperforms baselines under several few-shot settings. |
Massive End-to-end Speech Recognition Models with Time Reduction (2024.naacl-long)
Copied to clipboard
Weiran Wang, Rohit Prabhavalkar, Haozhe Shan, Zhong Meng, Dongseong Hwang, Qiujia Li, Khe Chai Sim, Bo Li, James Qin, Xingyu Cai, Adam Stooke, Chengjian Zheng, Yanzhang He, Tara Sainath, Pedro Moreno Mengibar
| Challenge: | Using the neural architecture of Google’s universal speech model, we reduce the frame rate and speed up training and inference. |
| Approach: | They propose to use the neural architecture of Google’s universal speech model with additional funnel pooling layers to significantly reduce the frame rate and speed up training and inference. |
| Outcome: | The proposed methods work with both connectionist temporal classification (CTC) and RNN-Transducer (RNN-T) and over two domains. |
DRS: Deep Question Reformulation With Structured Output (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing models like GPT-3 and Instruct-GPT lack the ability to reformulate unanswerable questions. |
| Approach: | They propose a zero-shot method that combines the strengths of LLMs with a DFS-based algorithm to iteratively explore potential entity combinations and constrain outputs using predefined entities. |
| Outcome: | The proposed method outperforms all baselines, including the GPT-3.5 model, on the unanswerable question reformulation task. |
HAT: Hardware-Aware Transformers for Efficient Natural Language Processing (2020.acl-main)
Copied to clipboard
| Challenge: | Extensive experiments on four machine translation tasks demonstrate that HAT can discover efficient models for different hardware (CPU, GPU, IoT device). |
| Approach: | They propose to construct a large design space with arbitrary encoder-decoder attention and heterogeneous layers and then train a SuperTransformer that efficiently produces many SubTransformers with weight sharing. |
| Outcome: | The proposed framework can find efficient models for different hardware (CPU, GPU, IoT device) it achieves 3 speedup, 3.7 smaller size over baseline Transformer; 2.7 speed up, 3.6 smaller sizes over Evolved Transformer with 12,041 less search cost and no performance loss. |
CoELM: Construction-Enhanced Language Modeling (2024.acl-long)
Copied to clipboard
| Challenge: | Recent studies show that integrating constructional information can improve the performance of pre-trained language models. |
| Approach: | They propose a construction-Enhanced language model that embeds constructional semantics into language models for natural language generation. |
| Outcome: | The proposed model outperforms existing models on various benchmarks. |
Beyond Demonstrations: Dynamic Vector Construction from Latent Representations (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for In-Context Learning (ICL) are sensitive to ICL-specific factors and rely on heuristic-based injection positions. |
| Approach: | They propose a method that extracts task-relevant representations from large language models and reinjects them during inference. |
| Outcome: | The proposed method outperforms few-shot In-Context Learning (ICL) and LoRA methods without repeated demonstration processing. |
CoreCodeBench: Decoupling Code Intelligence via Fine-Grained Repository-Level Tasks (2026.acl-long)
Copied to clipboard
Lingyue Fu, Hao Guan, Bolun Zhang, Haowei Yuan, Yaoming Zhu, Lin Qiu, ZongYu Wang, Xuezhi Cao, Xunliang Cai, Weiwen Liu, Weinan Zhang, Yong Yu
| Challenge: | Existing large language models for software engineering rely on coarse-grained pass rates obscuring specific cognitive bottlenecks. |
| Approach: | They propose a repository-level benchmark that dissects coding capabilities through atomized tasks. |
| Outcome: | The proposed framework achieves a 78.55% validity yield, surpassing the 31.7% retention rate of SWE-bench-Verified. |
Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers (2026.findings-acl)
Copied to clipboard
Zhiyang Zhang, Junda She, Kuo Cai, Bo Chen, Shiyao Wang, Xinchen Luo, Qiang Luo, Ruiming Tang, Han Li, Kun Gai, Guorui Zhou
| Challenge: | Existing methods for constructing item identifiers face bottlenecks due to their large output space and expensive vocabulary expansion and alignment training. |
| Approach: | They propose to use Large Language Models to develop general-purpose, semantically-aware recommender systems that can be generalized and reusable. |
| Outcome: | Experiments on real-world datasets show that GRAM outperforms baselines and significantly outperformed baselines. |
Conflict-Aware Memory for Embodied Agents: Enhancing Vector Data Quality via Detection Rules (2026.acl-long)
Copied to clipboard
| Challenge: | Embodied agents have successfully leveraged large language models (LLMs) to better transform human instructions and images into executable task plans. |
| Approach: | They propose Conflict Detection Rules to identify and manage data quality issues in vector knowledge bases and correct the index structure. |
| Outcome: | Experimental results show that planners with Conflict Detection Rules exceed the basic LLM planner by 15.25% and 14.25% in grammatical accuracy (GA) and interpretation accuracy (IA) on average. |
STAND-Guard: A Small Task-Adaptive Content Moderation Model (2025.coling-industry)
Copied to clipboard
| Challenge: | Content moderation is important for developing welcoming online platforms and responsible large language models. |
| Approach: | They propose a small task-adaptive coNtent moDeration model that can be easily adapted to new or customized content moderation tasks without extensive model tuning. |
| Outcome: | The proposed model is comparable to GPT-3.5-Turbo on unseen English binary classification tasks. |
What Makes for Good Visual Instructions? Synthesizing Complex Visual Reasoning Instructions for Visual Instruction Tuning (2025.coling-main)
Copied to clipboard
Yifan Du, Hangyu Guo, Kun Zhou, Wayne Xin Zhao, Jinpeng Wang, Chuyuan Wang, Mingchen Cai, Ruihua Song, Ji-Rong Wen
| Challenge: | Experimental results show that visual instruction tuning improves performance of Multi-modal Large Language Models (MLLMs) to extend the application scope of Large Language Modells, a surge of work augments LLMs with vision encoders to endow the ability of multi-modal cognition and reasoning. |
| Approach: | They propose a systematic approach to create high-quality visual reasoning instructions using a synthesize-complicate-reformulate paradigm. |
| Outcome: | The proposed method improves performance of MLLMs by 27.86% and 27.60% on MME-Perception and MME Cognition. |
Dialogue Response Selection with Hierarchical Curriculum Learning (2021.acl-long)
Copied to clipboard
Yixuan Su, Deng Cai, Qingyu Zhou, Zibo Lin, Simon Baker, Yunbo Cao, Shuming Shi, Nigel Collier, Yan Wang
| Challenge: | Empirical studies on three benchmark datasets with three state-of-the-art matching models demonstrate that the proposed learning framework significantly improves the model performance across various evaluation metrics. |
| Approach: | They propose a hierarchical curriculum learning framework that trains matching models in an “easy-to-difficult” scheme. |
| Outcome: | The proposed framework significantly improves the model performance across evaluation metrics on three benchmark datasets with three state-of-the-art matching models. |
Safety-Utility Conflicts Are Not Global: Surgical Alignment via Head-Level Diagnosis (2026.acl-long)
Copied to clipboard
| Challenge: | Existing mitigation strategies rely on global gradient geometry to resolve alignment conflicts . however, they overlook Modular Heterogeneity within Transformers, resulting in suboptimal trade-offs . Conflict-Aware Sparse Tuning (CAST) combines head-level diagnosis with sparse fine-tuning . |
| Approach: | They propose a framework that integrates head-level diagnosis with sparse fine-tuning to address this limitation. |
| Outcome: | The proposed framework integrates head-level diagnosis with sparse fine-tuning to reduce alignment conflicts in LLMs. |
AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts (2026.acl-long)
Copied to clipboard
Keyu Li, Junhao Shi, Yang Xiao, Mohan Jiang, Jie Sun, Yunze Wu, Dayuan Fu, Shijie Xia, Xiaojie Cai, Tianze Xu, Weiye Si, Wenjie Li, Dequan Wang, Pengfei Liu
| Challenge: | Existing benchmarks focus on single agentic capability, failing to capture long-horizon real-world scenarios. |
| Approach: | They propose a benchmark that evaluates 6 agentic capabilities across 32 real-world scenarios. |
| Outcome: | Experiments show that closed-source models outperform open-source model (48.4% vs 32.1%) integrating models with advanced scaffolds to form autonomous agents is a paradigm shift. |
DrDiff: Dynamic Routing Diffusion with Hierarchical Attention for Breaking the Efficiency-Quality Trade-off (2025.emnlp-main)
Copied to clipboard
Jusheng Zhang, Yijia Fan, Kaitong Cai, Zimeng Huang, Xiaofei Sun, Jian Wang, Chengpei Tang, Keze Wang
| Challenge: | et al., 2019; Brown e.t al, 2023; Touvron e t al; 2024; OpenAI, 2024) Large Language Models (LLMs) have demonstrated remarkable capabilities in knowledge encoding and contextual understanding during their pretraining phase. |
| Approach: | They propose a dynamic expert scheduling mechanism that allocates computational resources based on text complexity and a hierarchical sparse attention mechanism that adjusts attention patterns according to a variety of input lengths. |
| Outcome: | The proposed framework overpowers existing methods on long-text generation benchmarks. |
Mitigating Coordinate Prediction Bias from Positional Encoding Failures (2026.findings-acl)
Copied to clipboard
| Challenge: | Multimodal Large Language Models (MLLMs) excel at general vision-language tasks, but precise coordinate prediction remains a challenge. |
| Approach: | They propose a training-free, inference-time correction method to correct VPEs . they isolate position-unconditioned tendencies by shuffling VPE and use it to steer digit decoding . |
| Outcome: | The proposed method is training-free, inference-time correction method . it effectively rectifies coordinate drift, yielding consistent improvements without retraining . |
OneRec-Think: In-Text Reasoning for Generative Recommendation (2026.acl-long)
Copied to clipboard
Zhanyu Liu, Shiyao Wang, Xingmei Wang, Rongzhou Zhang, Jiaxin Deng, Honghui Bao, Jinghao Zhang, Wuchao Li, PengFei Zheng, Xiangyu Wu, Yifei Hu, Qigen Hu, Xinchen Luo, Lejian Ren, Zhang Zixing, Qianqian Wang, Kuo Cai, Yunfan Wu, Hongtao Cheng, Zexuan Cheng, Lu Ren, Huanjie Wang, Yi Su, Ruiming Tang, Kun Gai, Guorui Zhou
| Challenge: | Existing generative models lack the capacity for explicit and controllable reasoning, a key advantage of LLMs. |
| Approach: | They propose a framework that integrates dialogue, reasoning, and personalized recommendation. |
| Outcome: | Experiments across public benchmarks show state-of-the-art performance. |
SampleMix: A Sample-wise Pre-training Data Mixing Strategy by Coordinating Data Quality and Diversity (2025.findings-emnlp)
Copied to clipboard
Xiangyu Xi, Deyang Kong, Jian Yang, Jiawei Yang, Zhengyu Chen, Wei Wang, Jingang Wang, Xunliang Cai, Shikun Zhang, Wei Ye
| Challenge: | Existing methods for pretraining data mixing for large language models neglect significant inter-domain overlaps and commonalities, failing to control the global diversity of the constructed training dataset. |
| Approach: | They propose a sample-wise data mixture approach that performs global cross-domain sampling by systematically evaluating the quality and diversity of each sample. |
| Outcome: | The proposed method exceeds existing domain-based methods in multiple downstream tasks and perplexity assessments. |
Harmonizing Dense and Sparse Signals in Multi-turn RL: Dual-Horizon Credit Assignment for Industrial Sales Agents (2026.acl-industry)
Copied to clipboard
| Challenge: | Large language models for industrial sales require balancing long-term commercial objectives with immediate linguistic constraints such as fluency and compliance. |
| Approach: | They propose a framework that disentangles optimization across time scales by normalizing advantages from turn-level and session-level rewards before fusion. |
| Outcome: | The proposed framework outperforms the state-of-the-art GRPO model in conversion rate and identity detection rate. |
TextFlint: Unified Multilingual Robustness Evaluation Toolkit for Natural Language Processing (2021.acl-demo)
Copied to clipboard
Xiao Wang, Qin Liu, Tao Gui, Qi Zhang, Yicheng Zou, Xin Zhou, Jiacheng Ye, Yongxin Zhang, Rui Zheng, Zexiong Pang, Qinzhuo Wu, Zhengyan Li, Chong Zhang, Ruotian Ma, Zichu Fei, Ruijian Cai, Jun Zhao, Xingwu Hu, Zhiheng Yan, Yiding Tan, Yuan Hu, Qiyuan Bian, Zhihua Liu, Shan Qin, Bolin Zhu, Xiaoyu Xing, Jinlan Fu, Yue Zhang, Minlong Peng, Xiaoqing Zheng, Yaqian Zhou, Zhongyu Wei, Xipeng Qiu, Xuanjing Huang
| Challenge: | Existing approaches to textual robustness evaluation focus on slightly modifying the input data, which maintains the original meaning and results in a different prediction. |
| Approach: | They propose a multilingual robustness evaluation toolkit for NLP that integrates universal text transformations, task-specific transformations and adversarial attack. |
| Outcome: | The toolkit includes universal text transformation, task-specific transformation, adversarial attack, subpopulation, and their combinations to provide comprehensive robustness analyses. |
Query Enhanced Knowledge-Intensive Conversation via Unsupervised Joint Modeling (2023.acl-long)
Copied to clipboard
| Challenge: | Existing methods to retrieve knowledge-intensive conversations are based on external resources such as Wikipedia databases or search engine results. |
| Approach: | They propose an unsupervised query enhanced approach for knowledge-intensive conversations . they conduct experiments on three knowledge- intensive conversation datasets . |
| Outcome: | The proposed approach performs better than all unsupervised methods across three datasets and achieves competitive performance compared to supervised methods. |
TinyJudge: Unverifiable Constraint Alignment via Lightweight Specialist Ensembles (2026.acl-long)
Copied to clipboard
Yirong Zeng, Yufei Liu, Xiao Ding, Yutai Hou, Yuxian Wang, Wu Ning, Haonan Song, Dandan Tu, Qixun Zhang, Yuxiang He, Bibo Cai, Ting Liu
| Challenge: | Instruction Following (IF) is a core capability of LLMs, requiring strict adherence to diverse constraints. |
| Approach: | They propose a framework that uses tiny language models to evaluate instruction following . they propose to use a set of specialized tiny language model to provide rewards for soft constraints. |
| Outcome: | The proposed framework outperforms baseline models by 12% and speeds up training time by 3. |
Improving Event Definition Following For Zero-Shot Event Detection (2024.acl-long)
Copied to clipboard
Zefan Cai, Po-Nien Kung, Ashima Suvarna, Mingyu Ma, Hritik Bansal, Baobao Chang, P. Jeffrey Brantingham, Wei Wang, Nanyun Peng
| Challenge: | Existing approaches on zero-shot event detection train models on datasets annotated with known event types and prompt them with unseen event definitions. |
| Approach: | They propose to train models to better follow event definitions by using an automatic generated Diverse Event Definition dataset. |
| Outcome: | The proposed model outperforms existing models on three open benchmarks on zero-shot event detection. |
Large Language Models Meet Harry Potter: A Dataset for Aligning Dialogue Agents with Characters (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing models that can create open-domain dialogue agents lack character representation and annotations. |
| Approach: | They propose a dataset to study character alignment and character representation . it includes all dialogue sessions from the Harry Potter series and includes annotations . |
| Outcome: | The proposed dataset can be used as a universal benchmark for character-driven LLMs. |
QCRD: Quality-guided Contrastive Rationale Distillation for Large Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Recent research has focused on smaller, task-specific models enhanced by distilling knowledge from LLMs, but the diversity and quality of negative knowledge remains understudied. |
| Approach: | They propose a quality-guided contrastive rationale distillation framework that aims to enhance reasoning capabilities through contrastive knowledge learning. |
| Outcome: | The proposed method consistently outperforms existing distillation techniques yielding higher-quality rationales. |
RATION: Entropy-Driven Task-Adaptive Visual Attention Allocation Framework for Multimodal Reasoning (2026.findings-acl)
Copied to clipboard
| Challenge: | Prior studies have focused on strengthening multimodal reasoning by improving representation alignment or increasing computation, but these methods do not characterize the differences in visual demands across tasks. |
| Approach: | They propose an entropy-driven task-adaptive visual attention allocation framework that uses visual attention entropic as a control signal to dynamically allocate attention according to task demands. |
| Outcome: | The proposed framework achieves consistent performance gains across diverse reasoning tasks, datasets, and models, providing a clear direction toward more reliable multimodal reasoning. |
Omni-R1: Towards the Unified Generative Paradigm for Multimodal Reasoning (2026.findings-acl)
Copied to clipboard
| Challenge: | Early approaches focus on text-based reasoning, but they often follow a single task-specific reasoning pattern. |
| Approach: | They propose a generative multimodal reasoning paradigm that unifies diverse reasoning skills by generating intermediate images during the reasoning process. |
| Outcome: | The proposed model unifies diverse multimodal reasoning skills by generating intermediate images during the reasoning process. |
LANG: Reinforcement Learning for Multilingual Reasoning with Language-Adaptive Hint Guidance (2026.acl-long)
Copied to clipboard
Yuchun Fan, Bei Li, Peiguang Li, Yilin Wang, Yongyu Mu, Jian Yang, Xin Chen, Rongxiang Weng, Jingang Wang, Xunliang Cai, JingBo Zhu, Tong Xiao
| Challenge: | Existing methods for enhancing multi-step reasoning have not fully translated to multilingual contexts. |
| Approach: | They propose a framework that leverages language-conditioned hints to guide exploration in non-English reasoning tasks. |
| Outcome: | Empirical results show that the proposed framework improves reasoning performance without compromising language consistency. |
VideoStir: Understanding Long Videos via Spatio-Temporally Structured and Intent-Aware RAG (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for retrieval-augmented generation (RAG) to long videos are limited by limited context windows and flatten videos into independent segments. |
| Approach: | They propose a structured and intent-aware long-video RAG framework that structures a video as a spatio-temporal graph and then performs multi-hop retrieval to aggregate evidence across distant yet contextually related events. |
| Outcome: | The proposed framework is competitive with state-of-the-art baselines without auxiliary information. |
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. |