Papers by Pan Lu
AGD: Adversarial Game Defense Against Jailbreak Attacks in Large Language Models (2025.acl-long)
Copied to clipboard
Shilong Pan, Zhiliang Tian, Zhen Huang, Wanlong Yu, Zhihua Wen, Xinwang Liu, Kai Lu, Minlie Huang, Dongsheng Li
| Challenge: | Existing defenses, including post-training alignment and prompt engineering, struggle with adaptability to out-of-distribution (OOD) attacks. |
| Approach: | They propose an adversarial game-based defense method that dynamically adjusts LLMs’ internal representations to achieve a balanced trade-off between helpfulness and harmlessness. |
| Outcome: | The proposed method improves LLMs’ safety over all baselines. |
MathGenie: Generating Synthetic Data with Question Back-translation for Enhancing Mathematical Reasoning of LLMs (2024.acl-long)
Copied to clipboard
| Challenge: | Existing models have demonstrated outstanding capabilities in mathematical reasoning, but there is a performance gap between open-source models and closed-source ones. |
| Approach: | They propose a method for generating diverse and reliable math problems by leveraging the ground-truth solutions of the seed data. |
| Outcome: | The proposed model outperforms open-source models across five representative mathematical reasoning datasets. |
SafeAgent: Safeguarding LLM Agents via an Automated Risk Simulator (2026.acl-long)
Copied to clipboard
Xueyang Zhou, Weidong Wang, Lin Lu, Jiawen Shi, Guiyao Tie, Xu Yongtian, Lixing Chen, Pan Zhou, Neil Zhenqiang Gong, Lichao Sun
| Challenge: | SafeAgent improves agent safety through fully automated synthetic data generation. |
| Approach: | They propose a framework that improves agent safety through fully automated synthetic data generation. |
| Outcome: | The proposed framework outperforms closed-source models on two safety benchmarks and one real-world task. |
SocAoG: Incremental Graph Parsing for Social Relation Inference in Dialogues (2021.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on identifying entities' relations from the semantics of dialogues-they utilize either the attention mechanism or a refined token graph to locate informative words. |
| Approach: | They propose a sequential structure prediction task to incrementally parse SocAoG for dynamic inference upon any incoming utterance. |
| Outcome: | Empirical results show that the proposed model infers social relations more accurately than the state-of-the-art methods. |
mCLIP: Multilingual CLIP via Cross-lingual Transfer (2023.acl-long)
Copied to clipboard
Guanhua Chen, Lu Hou, Yun Chen, Wenliang Dai, Lifeng Shang, Xin Jiang, Qun Liu, Jia Pan, Wenping Wang
| Challenge: | Existing multilingual vision-language pretrained models are biased towards English due to the lack of sufficient non-English image-text pairs. |
| Approach: | They propose to train a retrieval-efficient dual-stream multilingual VLP model by aligning CLIP model and a multilingual text encoder through a novel Triangle Cross-modal Knowledge Distillation method. |
| Outcome: | Empirical results show that mCLIP achieves new state-of-the-art performance for both zero-shot and finetuned multilingual image-text retrieval tasks. |
Fact-Checking Complex Claims with Program-Guided Reasoning (2023.acl-long)
Copied to clipboard
| Challenge: | Fact-checking real-world claims often requires collecting multiple pieces of evidence and complex multi-step reasoning. |
| Approach: | They propose a novel fact-checking model that decomposes complex claims into simpler sub-tasks that can be solved using a shared library of specialized functions. |
| Outcome: | The proposed model outperforms seven baselines on two fact-checking datasets and has explicit output programs that benefit human debugging. |
SDiaReward: Modeling and Benchmarking Spoken Dialogue Rewards with Modality and Colloquialness (2026.acl-long)
Copied to clipboard
Jingyu Lu, Yuhan Wang, Fan Zhuo, Xize Cheng, Changhao Pan, Xueyi Pu, Yifu Chen, Chenyuhao Wen, Tianle Liang, Zhou Zhao
| Challenge: | SDiaReward is an end-to-end spoken dialogue system that integrates paralinguistic nuances and spontaneous nature of human conversation. |
| Approach: | They propose an end-to-end multi-turn reward model trained on SDiaReward-Dataset . it is a collection of episode-level preference pairs targeting modality and colloquiality gaps . |
| Outcome: | The proposed model outperforms general-purpose audio LLMs in episode-level evaluation. |
The Side Effects of Being Smart: Safety Risks in MLLMs’ Multi-Image Reasoning (2026.acl-long)
Copied to clipboard
Renmiao Chen, Yida Lu, Shiyao Cui, Xuan Ouyang, Victor Shea-Jay Huang, Shumin Zhang, Chengwei Pan, Han Qiu, Minlie Huang
| Challenge: | Recent advances in multimodal reasoning may pose new safety risks . evaluators neglect reasoningbased safety, where harm emerges only through MLLMs . |
| Approach: | They introduce a benchmark for multi-image reasoning safety that includes 2,676 instances . they find that models with more advanced multi- image reasoning are more vulnerable . |
| Outcome: | The proposed benchmark consists of 2,676 instances covering 9 multi-image relations . the results show that models with more advanced multi- image reasoning are more vulnerable . |
MathCoder-VL: Bridging Vision and Code for Enhanced Multimodal Mathematical Reasoning (2025.findings-acl)
Copied to clipboard
Ke Wang, Junting Pan, Linda Wei, Aojun Zhou, Weikang Shi, Zimu Lu, Han Xiao, Yunqiao Yang, Houxing Ren, Mingjie Zhan, Hongsheng Li
| Challenge: | Large Language Models (LMMs) struggle with simple tasks such as geometry, e.g., arithmetic, and reasoning. |
| Approach: | They propose to leverage code as supervision for cross-modal alignment . they propose to use FigCodifier and ImgCode-8.6M to synthesize novel mathematical figures . |
| Outcome: | The proposed model surpasses GPT-4o and Claude 3.5 Sonnet in the geometry problem-solving subset of MathVista, achieving improvements of 8.9% and 9.2%. |
HELENE: Hessian Layer-wise Clipping and Gradient Annealing for Accelerating Fine-tuning LLM with Zeroth-order Optimization (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) face memory challenges due to the high cost of backpropagation. |
| Approach: | They propose a zeroth-order (ZO) optimization that matches memory usage to inference . they propose scalable and memory-efficient zeroth order (ZE) optimizer that integrates annealed A-GNB gradients with diagonal Hessian estimation and layer-wise clipping as a second-order pre-conditioner. |
| Outcome: | The proposed algorithm outperforms state-of-the-art methods with an average speedup of 20 over MeZO on RoBERTa-large and OPT-1.3B. |
Virtual Context Enhancing Jailbreak Attacks with Special Token Injection (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Existing jailbreak attacks target the two phases of user interaction: prompt input and model computation. |
| Approach: | They propose a new tool that leverages special tokens to improve jailbreak attacks . they found that the tool can increase success rates of existing jailbreak methods by 40% . |
| Outcome: | The proposed solution can improve success rates of four widely used jailbreak methods by approximately 40% across various LLMs. |
TheoremQA: A Theorem-driven Question Answering Dataset (2023.emnlp-main)
Copied to clipboard
| Challenge: | Recent LLMs like GPT-4 and PaLM-2 have made tremendous progress in solving fundamental math problems like GSM8K by achieving over 90% accuracy. |
| Approach: | They propose to use theorem-driven question-answering dataset to evaluate AI models' ability to apply theoretic concepts to solving challenging science problems. |
| Outcome: | TheoremQA is curated by domain experts and contains 800 high-quality questions covering 350 theoremics from Math, Physics, EE&CS, and Finance. |
VDebugger: Harnessing Execution Feedback for Debugging Visual Programs (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Visual programs are executable code generated by large language models to address visual reasoning problems. |
| Approach: | They propose a critic-refiner framework that localizes and debugs visual programs by tracking execution step by step. |
| Outcome: | The proposed framework detects and corrects program errors leveraging detailed execution feedback, improving interpretability and accuracy. |
D-Artemis: A Deliberative Cognitive Framework for Mobile GUI Multi-Agents (2026.findings-acl)
Copied to clipboard
Hongze Mi, Yibo Feng, WenJie Lu, Yuqi Wang, Jinyuan Li, Song Cao, He Cui, Tengfei Tian, Xuelin Zhang, Haotian Luo, Di Sun, Jun Fang, Hua Chai, Naiqiang Tan, Gang Pan
| Challenge: | Graphical User Interface (GUI) agents aim to automate a wide spectrum of human tasks by emulating user interaction. |
| Approach: | They propose a deliberative framework that leverages a fine-grained tip retrieval mechanism to inform its decision-making process. |
| Outcome: | The proposed framework achieves SOTA among open-source general models on AndroidWorld and ScreenSpot-V2 . it leverages a fine-grained, app-specific tip retrieval mechanism to inform its decision-making process . |
Twitter Homophily: Network Based Prediction of User’s Occupation (P19-1)
Copied to clipboard
| Challenge: | Existing approaches to predicting Twitter users' demographic attributes exploit, select, and combine various features generated from text and network to achieve the best performance. |
| Approach: | They extend existing Twitter occupational class prediction data set and exploit social network homophily to achieve competitive performance. |
| Outcome: | The proposed method achieves better performance on a dataset with a small fraction of the training data. |
ELISA-EDL: A Cross-lingual Entity Extraction, Linking and Localization System (N18-5)
Copied to clipboard
| Challenge: | ELISA-EDL is a cross-lingual entity extraction, linking and localization system for Wikipedia languages. |
| Approach: | They propose a cross-lingual entity extraction, linking and localization system for English speakers . it extracts entities from unstructured text in any of 282 Wikipedia languages and links them to English knowledge bases . |
| Outcome: | The proposed system extracts entity mentions from Wikipedia and links them to English knowledge bases and visualizes locations related to disaster topics on a world heatmap. |
QACheck: A Demonstration System for Question-Guided Multi-Hop Fact-Checking (2023.emnlp-demo)
Copied to clipboard
| Challenge: | Existing fact-checking systems lack transparency in their decision-making process, making it difficult for users to comprehend their reasoning process. |
| Approach: | They propose a Question-guided Multi-hop Fact-Checking system which asks a series of questions critical for verifying a claim. |
| Outcome: | The proposed model provides a comprehensive report detailing its reasoning process, guided by a sequence of questions, answer pairs, and the source of evidence supporting each question. |
Merger-as-a-Stealer: Stealing Targeted PII from Aligned LLMs with Model Merging (2025.emnlp-main)
Copied to clipboard
| Challenge: | Model merging is a promising approach for updating large language models . but unmonitored mergers can introduce significant security vulnerabilities . |
| Approach: | They propose a model merging attack surface where a malicious merger can extract PII from an aligned model with model merg. |
| Outcome: | The proposed framework can extract PII from an aligned model with model merging. |
OctoTools: A Multi-Agent Framework with Extensible Tools for Complex Reasoning (2026.acl-long)
Copied to clipboard
| Challenge: | Existing prompting methods for large language models (LLMs) are restricted to specialized domains, limited tool types, or require additional training data. |
| Approach: | They propose a training-free, user-friendly, and easily extensible multi-agent framework designed to tackle complex reasoning across diverse domains. |
| Outcome: | The proposed framework outperforms AutoGen, GPT-Functions, and LangChain by up to 10.6% when given the same set of tools. |
NOVA: An Iterative Planning Framework for Enhancing Scientific Innovation with Large Language Models (2025.findings-acl)
Copied to clipboard
Xiang Hu, Hongyu Fu, Jinge Wang, Yifeng Wang, Zhikun Li, Renjun Xu, Yu Lu, Yaochu Jin, Lili Pan, Zhenzhong Lan
| Challenge: | Existing approaches to generate research ideas rely on retrieval or prompt engineering to generate ideas. |
| Approach: | They propose a method that uses iterative planning and search to boost creative potential of LLMs by integrating external knowledge with broader and deeper insights. |
| Outcome: | The proposed method outperforms the current state-of-the-art in generating 2.5 times more top-rated ideas based on 170 seed papers in a Swiss Tournament evaluation. |
Alignment for Efficient Tool Calling of Large Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Recent advances in tool learning have enabled large language models to integrate external tools, enhancing their task performance by expanding their knowledge boundaries. |
| Approach: | They propose a framework that combines probabilistic knowledge boundary estimation with dynamic decision-making to allow LLMs to better assess when to invoke tools based on their confidence. |
| Outcome: | The proposed framework shows significant improvements in tool efficiency by reducing unnecessary tool usage. |
WaterSeeker: Pioneering Efficient Detection of Watermarked Segments in Large Documents (2025.findings-naacl)
Copied to clipboard
| Challenge: | Existing methods focus on distinguishing fully watermarked text from non-watermarked text, overlooking real-world scenarios where LLMs generate only brief segments within longer documents. |
| Approach: | They propose a method to detect watermarked segments in large documents using an anomaly extraction method and a local traversal. |
| Outcome: | The proposed method achieves a superior balance between detection accuracy and computational efficiency. |
Event Extraction as Multi-turn Question Answering (2020.findings-emnlp)
Copied to clipboard
| Challenge: | Current approaches to event extraction fail to model rich interactions among event types and arguments of different roles. |
| Approach: | They propose a new paradigm that formulates event extraction as multi-turn question answering . they propose to use reading comprehension problems to extract triggers and arguments . |
| Outcome: | The proposed approach outperforms current state-of-the-art on argument extraction tasks . it makes full use of dependency among arguments and event types, and generalizes well . |
CoVerRL: Breaking the Consensus Trap in Label-Free Reasoning via Generator-Verifier Co-Evolution (2026.acl-long)
Copied to clipboard
Teng Pan, Yuchen Yan, Zixuan Wang, Ruiqing Zhang, Guiyang Hou, Wenqi Zhang, Weiming Lu, Jun Xiao, Yongliang Shen
| Challenge: | Label-free reinforcement learning enables large language models to improve reasoning capabilities . but as training maximizes self-consistency, output diversity collapses, authors say . authors propose a framework where a single model alternates between generator and verifier roles . |
| Approach: | They propose a framework where a model alternates between generator and verifier roles, bootstrapping each other. |
| Outcome: | Experiments show that CoVerRL outperforms label-free baselines on reasoning benchmarks . the framework can be used to improve reasoning abilities without ground-truth supervision . |
Probability-Consistent Preference Optimization for Enhanced LLM Reasoning (2025.findings-acl)
Copied to clipboard
Yunqiao Yang, Houxing Ren, Zimu Lu, Ke Wang, Weikang Shi, Aojun Zhou, Junting Pan, Mingjie Zhan, Hongsheng Li
| Challenge: | Recent advances in preference optimization have demonstrated significant potential for improving mathematical reasoning capabilities in large language models. |
| Approach: | They propose a framework that establishes two quantitative metrics for preference selection: surface-level answer correctness and intrinsic token-level probability consistency. |
| Outcome: | The proposed framework outperforms existing outcome-only criterion approaches across a diverse range of LLMs and benchmarks. |
A Survey on Training-free Alignment of Large Language Models (2025.findings-emnlp)
Copied to clipboard
Birong Pan, Yongqi Li, Weiyu Zhang, Wenpeng Lu, Mayi Xu, Shen Zhou, Yuanyuan Zhu, Ming Zhong, Tieyun Qian
| Challenge: | a survey of large language models (LLMs) aims to ensure outputs adhere to human values, ethical standards, and legal norms. |
| Approach: | They present the first systematic review of TF alignment methods . they categorize them by stages of pre-decoding, in-decoder and post-decoration . |
| Outcome: | The proposed methods are based on training-free (TF) alignment techniques . they are able to be used in open-source and closed-source environments without retraining . |
Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning (2021.acl-long)
Copied to clipboard
| Challenge: | Existing methods for solving geometric problems are either small in scale or not publicly available. |
| Approach: | They propose a large-scale benchmark for geometric problem solving using formal language and symbolic reasoning. |
| Outcome: | The proposed approach parses geometry problems into formal language and performs symbolic reasoning step by step. |
Comprehensive Benchmarking of Long-Form Speech Generation in Diverse Scenarios (2026.findings-acl)
Copied to clipboard
Changhao Pan, Rui Yang, Han Wang, Zhuan Zhou, Xuming He, Wenxiang Guo, Ziyue Jiang, Ruiqi Li, Yu Zhang, Chenyuhao Wen, Ke Lei, Xiang Yin, Jingyu Lu, Zhiyuan Zhu, Zhou Zhao
| Challenge: | Existing evaluation benchmarks for long-form speech are limited to limited domains, creating a significant gap with the diverse downstream applications. |
| Approach: | They propose a benchmark that decomposes "long-form speech quality" into specific, disentangled dimensions. |
| Outcome: | The proposed benchmark decomposes “long-form speech quality” into specific, disentangled dimensions. |
UniGeo: Unifying Geometry Logical Reasoning via Reformulating Mathematical Expression (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing work on geometry problem solving treats calculation and proving as two specific tasks hindering a deep model to unify reasoning ability on multiple math tasks. |
| Approach: | They propose a large-scale Unified Geometry problem benchmark to unify geometry on multiple math tasks. |
| Outcome: | The proposed framework outperforms the existing model with 5.6% and 3.2% accuracies on calculation and proving problems. |
Towards Universal Debiasing for Language Models-based Tabular Data Generation (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing large language models have exacerbated fairness issues in tabular data generation . inherent historical biases in tabulated data cause LLMs to exacerbate fairness problems . |
| Approach: | They propose a universal debiasing framework that minimizes group-level dependencies . it leverages the autoregressive structure and analytic sampling distributions of LLM-based tabular data generators . |
| Outcome: | The proposed framework minimizes group-level dependencies while reducing mutual information between advantaged and protected attributes. |
MASPO: Unifying Gradient Utilization, Probability Mass, and Signal Reliability for Robust and Sample-Efficient LLM Reasoning (2026.acl-long)
Copied to clipboard
Xiaoliang Fu, Jiaye Lin, Yangyi Fang, Binbin Zheng, Chaowen Hu, Zekai Shao, Cong Qin, Lu Pan, Ke Zeng, Xunliang Cai
| Challenge: | Existing RLVR algorithms rely on rigid, uniform, and symmetric trust region mechanisms . current algorithms lack robustness, asymmetric signal reliability and inefficient gradient utilization . |
| Approach: | They propose a framework to harmonize three dimensions of RLVR algorithms, a paper argues . a binary cutoff is used to discard valuable reinforcement signals, they argue . |
| Outcome: | The proposed framework outperforms baselines in evaluating a robust RLVR solution. |
Multimodal Procedural Planning via Dual Text-Image Prompting (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Embodied agents have demonstrated performance in following instructions informed by texts and images . however, the potential of models providing useful guidelines for humans to complete tasks remains underexplored . |
| Approach: | They propose a multimodal procedural planning task that generates paired text-image plans . this task provides more complementary and informative guidance than unimodal plans a . authors propose modality prompting methods that leverage zero-shot reasoning ability . |
| Outcome: | The proposed method improves the interaction in dual modalities and provides more information than unimodal plans. |
From log 𝜋 to 𝜋: Taming Divergence in Soft Clipping via Bilateral Decoupled Decay of Probability Gradient Weight (2026.acl-long)
Copied to clipboard
Xiaoliang Fu, Jiaye Lin, Yangyi Fang, Chaowen Hu, Cong Qin, Zekai Shao, Binbin Zheng, Lu Pan, Ke Zeng
| Challenge: | Standard algorithms for Large Language Models (LLMs) enforce stability via "hard clipping" but relying on log-probability gradient yields divergent weights as probabilities vanish, destabilizing LLM training. |
| Approach: | They propose a decoupled gradient policy optimization that uses a decay mechanism to decouple the probability of a boundary token. |
| Outcome: | The proposed algorithm outperforms baselines on various mathematical benchmarks. |
Can LLM Watermarks Robustly Prevent Unauthorized Knowledge Distillation? (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Model (LLM) watermarking is radioactive and enables the detection of watermarks inherited by student models when trained on the outputs of watermarked teacher models. |
| Approach: | They propose two types of watermark removal attacks that allow student models to perform untraceable knowledge distillation while avoiding watermark inheritance. |
| Outcome: | The proposed attacks eliminate inherited watermarks while maintaining knowledge transfer efficiency and low computational overhead. |
DaMo: Data Mixing Optimizer in Fine-tuning Multimodal LLMs for Mobile Phone Agents (2026.findings-acl)
Copied to clipboard
Kai Shi, Jun Yang, Ni Yang, Binqiang Pan, Qingsong Xie, null Zhangchao, Zhenyu Yang, Tianhuang Su, Haonan Lu
| Challenge: | Mobile Phone Agents (MPAs) have attracted huge attention due to their practicability in a multitude of scenarios. |
| Approach: | They propose a data mixture optimization solution that extrapolates optimal data mixtures from a trainable network. |
| Outcome: | The proposed model outperforms existing methods on open-source benchmarks and on open source benchmarks. |
Error Analysis of Uyghur Name Tagging: Language-specific Techniques and Remaining Challenges (L18-1)
Copied to clipboard
| Challenge: | despite efforts at name tagging, there is limited understanding on the performance ceiling . despite the high-resource language, there are very few natural language processing tools available . |
| Approach: | They propose to use a machine learning model to identify Uyghur name tagger errors . they conclude that such a model is unlikely to be effective for Uygur, or low-resource languages . |
| Outcome: | The proposed model is unlikely to be effective for Uyghur, or low-resource languages in general, the authors argue . they show that the proposed model can be used for high-res languages with superficial features . |
Simple or Complex? Complexity-controllable Question Generation with Soft Templates and Deep Mixture of Experts Model (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Existing work on complex questions does not consider controlling complexity of generated questions. |
| Approach: | They propose an end-to-end neural complexity-controllable question generation model that incorporates a mixture of experts as the selector of soft templates to capture question similarity while avoiding the expensive construction of actual templates. |
| Outcome: | The proposed model is superior to state-of-the-art methods in both automatic and manual evaluations on two benchmark QA datasets. |
LILA: A Unified Benchmark for Mathematical Reasoning (2022.emnlp-main)
Copied to clipboard
Swaroop Mishra, Matthew Finlayson, Pan Lu, Leonard Tang, Sean Welleck, Chitta Baral, Tanmay Rajpurohit, Oyvind Tafjord, Ashish Sabharwal, Peter Clark, Ashwin Kalyan
| Challenge: | Towards evaluating and improving AI systems in this domain, we propose a mathematical reasoning benchmark based on 23 diversetasks . |
| Approach: | They propose a mathematical reasoning benchmark that includes 23 diverse tasks . they extend the benchmark by collecting task instructions and solutions in the form of Python programs . |
| Outcome: | The proposed model improves on multi-tasking while the best performing model only achieves 60.40%. |
Streaming Hallucination Detection in Long Chain-of-Thought Reasoning (2026.findings-acl)
Copied to clipboard
Haolang Lu, Minghui Pan, Ripeng LI, Guoshun Nan, Jialin Zhuang, Zijie Zhao, Zhongxiang Sun, Kun Wang, Yang Liu
| Challenge: | Long chain-of-thought reasoning improves performance of large language models, yet hallucinations in such settings often emerge subtly and propagate across reasoning steps. |
| Approach: | They propose to treat step-level hallucination judgments as local observations and introduce a cumulative prefix-level signal that tracks the global evolution of the reasoning state over the entire trajectory. |
| Outcome: | The proposed method enables streaming hallucination detection in long CoT reasoning, providing real-time, interpretable evidence. |
Alignment with Fill-In-the-Middle for Enhancing Code Generation (2025.emnlp-main)
Copied to clipboard
Houxing Ren, Zimu Lu, Weikang Shi, Haotian Hou, Yunqiao Yang, Ke Wang, Aojun Zhou, Junting Pan, Mingjie Zhan, Hongsheng Li
| Challenge: | Existing methods for generating test cases with limited training data are not reliable and may be counterproductive. |
| Approach: | They propose a method that splits code snippets into smaller, granular blocks, creating more diverse DPO pairs from the same test cases. |
| Outcome: | The proposed approach shows significant improvements in code generation tasks on benchmark datasets such as HumanEval (+), MBPP (+), and APPS. |
A Survey of Deep Learning for Mathematical Reasoning (2023.acl-long)
Copied to clipboard
| Challenge: | a survey of deep learning for mathematical reasoning examines the field . a comprehensive reading list is provided to assist readers interested in the field. |
| Approach: | They present a survey of deep learning for mathematical reasoning over the past decade . they outline directions for future research and highlight potential for further exploration . |
| Outcome: | The proposed framework is based on the results of a decade-long survey of deep learning for mathematical reasoning. |
TART: An Open-Source Tool-Augmented Framework for Explainable Table-based Reasoning (2025.findings-naacl)
Copied to clipboard
| Challenge: | Current Large Language Models lack ability to understand table structures and apply precise numerical reasoning. |
| Approach: | They propose a tool-augmented reasoning framework for table-based tasks that integrates LLMs with specialized tools. |
| Outcome: | The proposed framework improves on the TOOLTAB dataset, a benchmark for LLMs in table–tool integration. |
Versatile Framework for Song Generation with Prompt-based Control (2025.findings-emnlp)
Copied to clipboard
Yu Zhang, Wenxiang Guo, Changhao Pan, Zhiyuan Zhu, Ruiqi Li, Jingyu Lu, Rongjie Huang, Ruiyuan Zhang, Zhiqing Hong, Ziyue Jiang, Zhou Zhao
| Challenge: | Existing methods for song generation fail to generate vocals with prompt-based control and proper alignment. |
| Approach: | VersBand is a multi-task song generation framework for synthesizing high-quality songs with prompt-based control. |
| Outcome: | Experimental results show that VersBand performs better than baseline models across multiple song generation tasks. |
MarkLLM: An Open-Source Toolkit for LLM Watermarking (2024.emnlp-demo)
Copied to clipboard
Leyi Pan, Aiwei Liu, Zhiwei He, Zitian Gao, Xuandong Zhao, Yijian Lu, Binglin Zhou, Shuliang Liu, Xuming Hu, Lijie Wen, Irwin King, Philip Yu
| Challenge: | Large Language Models (LLMs) embed imperceptible yet algorithmically detectable signals in outputs to identify LLM-generated text. |
| Approach: | They propose to develop an open-source toolkit for LLM watermarking that embeds imperceptible yet algorithmically detectable signals in model outputs to identify LLM-generated text. |
| Outcome: | MarkLLM provides a unified framework for implementing LLM watermarking algorithms, while providing user-friendly interfaces to ensure ease of access. |
Model Editing Harms General Abilities of Large Language Models: Regularization to the Rescue (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods that edit large language models with updated knowledge can cause side effects on the general abilities of LLMs such as reasoning, natural language inference, and question answering. |
| Approach: | They propose to regularize the edit update weights by imposing constraints on their complexity based on the RElative Change in weighT. |
| Outcome: | The proposed method can significantly mitigate the side effects while maintaining over 94% editing performance. |
Protein Large Language Models: A Comprehensive Survey (2025.findings-emnlp)
Copied to clipboard
Yijia Xiao, Wanjia Zhao, Junkai Zhang, Yiqiao Jin, Han Zhang, Zhicheng Ren, Renliang Sun, Haixin Wang, Guancheng Wan, Pan Lu, Xiao Luo, Yu Zhang, James Zou, Yizhou Sun, Wei Wang
| Challenge: | Existing studies focus on specific aspects or applications, but this study provides a comprehensive overview of Protein-specific large language models. |
| Approach: | This paper proposes a structured taxonomy of state-of-the-art ProteinLLMs . they analyze how they leverage large-scale protein sequence data for improved accuracy . |
| Outcome: | The proposed model covers their architectures, training datasets, evaluation metrics, and diverse applications. |
How to Allocate, How to Learn? Dynamic Rollout Allocation and Advantage Modulation for Policy Optimization (2026.findings-acl)
Copied to clipboard
Yangyi Fang, Jiaye Lin, Xiaoliang Fu, Cong Qin, Haolin Shi, Chaowen Hu, Lu Pan, Ke Zeng, Xunliang Cai
| Challenge: | Existing methods for reinforcement learning with verifiable rewards are limited by the complexity of the problem and the complexity. |
| Approach: | They propose a theoretically-grounded dual-pronged optimization framework for reinforcement learning with verifiable rewards that compensates for gradient attenuation of high-confidence correct actions while utilizing entropy changes as computable indicators to stabilize excessive update magnitudes. |
| Outcome: | The proposed framework compensates for gradient attenuation of high-confidence correct actions while utilizing entropy changes as computable indicators to stabilize excessive update magnitudes. |
SCITAB: A Challenging Benchmark for Compositional Reasoning and Claim Verification on Scientific Tables (2023.emnlp-main)
Copied to clipboard
| Challenge: | Current scientific fact-checking benchmarks exhibit several shortcomings, such as biases arising from crowd-sourced claims and an over-reliance on text-based evidence. |
| Approach: | They present a dataset of 1.2K expert-verified scientific claims that require compositional reasoning for verification. |
| Outcome: | The proposed model outperforms existing models in table-based pretraining models and large language models. |
Rethinking Stateful Tool Use in Multi-Turn Dialogues: Benchmarks and Challenges (2025.findings-acl)
Copied to clipboard
Hongru Wang, Wenyu Huang, Yufei Wang, Yuanhao Xi, Jianqiao Lu, Huan Zhang, Nan Hu, Zeming Liu, Jeff Z. Pan, Kam-Fai Wong
| Challenge: | Existing benchmarks that assess Language Models (LMs) as Language Agents (LAs) for tool use focus on stateless, single-turn interactions or partial evaluations, overlooking the inherent stateful nature of interactions in multi-turn applications. |
| Approach: | They propose a multi-turn dialogue dataset with stateful tool interactions considering the whole life cycle of tool use across six key tasks in three stages . they also build VirtualMobile – an embodied virtual mobile evaluation environment to simulate API calls and assess the robustness of the created APIs. |
| Outcome: | The proposed dataset evaluates 13 open- and closed-source LLMs and provides detailed analysis at each stage. |
Beyond Itinerary Planning—A Real-World Benchmark for Multi-Turn and Tool-Using Travel Tasks (2026.acl-long)
Copied to clipboard
| Challenge: | Existing studies on LLM performance on travel planning have shown that existing settings are limited due to limited domain coverage, insufficient modeling of users’ implicit preferences in multi-turn conversations, and a lack of evaluation of agents’ capability boundaries. |
| Approach: | They propose a benchmark to evaluate LLMs' planning and tool-use abilities in real-world settings by collecting user queries, user preferences, and tools from real scenarios. |
| Outcome: | The proposed benchmark evaluates agents' capabilities in real-world settings and shows that even advanced models exhibit imbalanced performance across different capabilities. |
PIVOINE: Instruction Tuning for Open-world Entity Profiling (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for information extraction focus on a closed-world setting, but PIVOINE is a promising solution to tackle the open-world problem of entity profiling. |
| Approach: | They propose to develop an LLM that performs Open-world Entity Profiling with instruction tuning to extract desirable entity profiles . they construct INSTRUCTOPENWIKI, a substantial instruction-tuning dataset for Open-World Entity Profiles . |
| Outcome: | The proposed model outperforms existing methods and ChatGPT-based baselines on unseen and out-of-ontology cases. |
AiraXiv: An AI-Driven Open-Access Platform for Human and AI Scientists (2026.acl-demo)
Copied to clipboard
| Challenge: | Recent advances in artificial intelligence (AI) have accelerated the growth of both human-authored and AI-generated research outputs. |
| Approach: | They propose an AI-driven open-access platform built on open preprints, AI-augmented analysis and review, and reader feedback. |
| Outcome: | The proposed platform supports human scientists through an interactive UI and AI scientists through Model Context Protocol (MCP)-based interactions. |
Stealing Training Data from Large Language Models in Decentralized Training through Activation Inversion Attack (2025.acl-long)
Copied to clipboard
| Challenge: | Decentralized training is a resource-efficient framework to democratize training of large language models. |
| Approach: | They propose an activation inversion attack to exploit privacy leakage from training data . they construct a shadow dataset comprising text labels and corresponding activations . |
| Outcome: | The proposed attack surface is based on a shadow dataset and public datasets . the proposed attack model reconstructs training data from activations in victim decentralized training. |
Are LLMs Capable of Data-based Statistical and Causal Reasoning? Benchmarking Advanced Quantitative Reasoning with Data (2024.findings-acl)
Copied to clipboard
| Challenge: | Quantitative reasoning with data is a critical skill to analyze data, yet the assessment of such ability remains limited. |
| Approach: | They propose a quantitative reasoning with data benchmark to evaluate Large Language Models' ability in statistical and causal reasoning with real-world data. |
| Outcome: | The proposed model GPT-4 achieves an accuracy of 58%, while open-source model Deepseek-coder-instruct gets the highest accuracy of 37%. |
Enhancing Open-Domain Task-Solving Capability of LLMs via Autonomous Tool Integration from GitHub (2025.acl-long)
Copied to clipboard
Bohan Lyu, Xin Cong, Heyang Yu, Pan Yang, Cheng Qian, Zihe Wang, Yujia Qin, Yining Ye, Yaxi Lu, Chen Qian, Zhong Zhang, Yukun Yan, Yankai Lin, Zhiyuan Liu, Maosong Sun
| Challenge: | Existing approaches lack flexibility to address diverse and ever-evolving user queries in open domains. |
| Approach: | They propose to evaluate LLMs on open-domain knowledge that requires tools to solve diverse and ever-evolving user queries. |
| Outcome: | The proposed system outperforms baselines in the open domain task-solving benchmark. |