Papers by Zhu Cao
Chinese SafetyQA: A Safety Short-form Factuality Benchmark for Large Language Models (2025.acl-long)
Copied to clipboard
Yingshui Tan, Boren Zheng, Baihui Zheng, Kerui Cao, Huiyun Jing, Jincheng Wei, Jiaheng Liu, Yancheng He, Wenbo Su, Xiaoyong Zhu, Bo Zheng, Kaifu Zhang
| Challenge: | Large language models have created significant safety concerns . factuality ability is crucial in determining whether they can be deployed and applied safely and compliantly within specific regions. |
| Approach: | They propose a benchmark to evaluate the factuality of large language models in China . they evaluate the models' ability to provide accurate and reliable information . |
| Outcome: | The proposed benchmark evaluates the factuality abilities of existing LLMs and compares them to LLM abilities. |
MRRL: Modifying the Reference via Reinforcement Learning for Non-Autoregressive Joint Multiple Intent Detection and Slot Filling (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing non-autoregressive models for multiple intent detection and slot filling have limited overall accuracy due to multi-modality problem and lack of alignment between correct predictions. |
| Approach: | They propose a method for multiple intent detection and slot filling that introduces a modifier and employs reinforcement learning to modify the reference. |
| Outcome: | The proposed method outperforms the previous best approach by 3.6 overall accuracy on MixATIS dataset. |
SILO-BENCH: A Scalable Environment for Evaluating Distributed Coordination in Multi-Agent LLM Systems (2026.acl-long)
Copied to clipboard
Yuzhe Zhang, Feiran Liu, Yi Shan, Xinyi Huang, Xin Yang, Yueqi Zhu, Xuxin Cheng, Cao Liu, Ke Zeng, Terry Jingchen Zhang, Wenyuan Jiang
| Challenge: | Existing benchmarks conflate coordination ability with role-based priors. |
| Approach: | They propose a role-free benchmark for evaluating free-form collaboration under information silos. |
| Outcome: | The proposed benchmark systematically probes coordination capabilities under information silos using 54 configurations and 3 frontier LLMs. |
ScaleBox: Enabling High-Fidelity and Scalable Code Verification for Large Language Models (2026.acl-demo)
Copied to clipboard
Jiasheng Zheng, Xin Zheng, Boxi Cao, Pengbo Wang, Zhengzhao Ma, Qiming Zhu, Jiazhen Jiang, Yaojie Lu, Hongyu Lin, Xianpei Han, Le Sun
| Challenge: | Existing code sandboxes fail to provide accurate verification and efficiency under high-concurrency workloads. |
| Approach: | They propose a high-fidelity code verification system that provides sandbox feedback for RL training and evaluation. |
| Outcome: | The proposed system outperforms heuristic-matching baselines on LiveCodeBench and training stability on high-concurrency workloads. |
Sharing, Teaching and Aligning: Knowledgeable Transfer Learning for Cross-Lingual Machine Reading Comprehension (2023.findings-emnlp)
Copied to clipboard
| Challenge: | X-STA is a new approach for cross-lingual machine reading comprehension . the variation of answer span positions in different languages makes it difficult to transfer knowledge across languages. |
| Approach: | They propose a method that leverages an attentive teacher to subtly transfer the answer spans of the source language to the answer output space of the target. |
| Outcome: | The proposed method outperforms state-of-the-art approaches on three multi-lingual datasets. |
Gaussian Process based Deep Dyna-Q approach for Dialogue Policy Learning (2021.findings-acl)
Copied to clipboard
| Challenge: | Reinforcement learning (RL) is the main dialogue policy learning method in recent years. |
| Approach: | They propose a Gaussian Process based Deep Dyna-Q approach to dialogue policy learning . they propose evaluating the quality of experiences generated by the world model using a discriminator . |
| Outcome: | The proposed approach improves the effectiveness and efficiency of dialogue policy learning by 20% with fewer human-machine interactions. |
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. |
RWKV: Reinventing RNNs for the Transformer Era (2023.findings-emnlp)
Copied to clipboard
Bo Peng, Eric Alcaide, Quentin Anthony, Alon Albalak, Samuel Arcadinho, Stella Biderman, Huanqi Cao, Xin Cheng, Michael Chung, Leon Derczynski, Xingjian Du, Matteo Grella, Kranthi Gv, Xuzheng He, Haowen Hou, Przemyslaw Kazienko, Jan Kocon, Jiaming Kong, Bartłomiej Koptyra, Hayden Lau, Jiaju Lin, Krishna Sri Ipsit Mantri, Ferdinand Mom, Atsushi Saito, Guangyu Song, Xiangru Tang, Johan Wind, Stanisław Woźniak, Zhenyuan Zhang, Qinghua Zhou, Jian Zhu, Rui-Jie Zhu
| Challenge: | recurrent neural networks struggle to match the performance of Transformers due to limitations in parallelization and scalability. |
| Approach: | They propose a model architecture that combines the efficient parallelizable training of transformers with the efficient inference of RNNs. |
| Outcome: | The proposed model performs on par with similarly sized RNNs, suggesting future work can leverage this architecture to create more efficient models. |
Format-Adapter: Improving Reasoning Capability of LLMs by Adapting Suitable Format (2026.findings-acl)
Copied to clipboard
Dingzirui Wang, Xuanliang Zhang, Rongyu Cao, Longxu Dou, Xianzhen Luo, Yingwei MA, Qingfu Zhu, Binhua Li, Fei Huang, Yongbin Li
| Challenge: | Prior work showed that multiple reasoning formats outperform a single format when generating multiple answers. |
| Approach: | They propose a method to measure reasoning error when generating multiple answers . they propose 'formatadapter' which generates and selects suitable reasoning formats . |
| Outcome: | The proposed method achieves a 4.3% performance improvement over previous works on math and commonsense reasoning tasks. |
Look Light, Think Heavy: What Multimodal Chain-of-Thought Reasoning Can and Cannot Do (2026.acl-long)
Copied to clipboard
| Challenge: | Existing open-source models often yield only marginal overall improvements, possibly due to an overemphasis on mathematical reasoning at the expense of broader capabilities. |
| Approach: | They evaluate 12 multimodal tasks using 14 non-reasoning models and 8 reasoning models. |
| Outcome: | The proposed method is effective in multimodal reasoning tasks, the authors show . they show that it lacks the ability to maintain deep visual introspection throughout the reasoning process. |
Reliably Bounding False Positives: A Zero-Shot Machine-Generated Text Detection Framework via Multiscaled Conformal Prediction (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods focus excessively on detection accuracy, neglecting the societal risks posed by high false positive rates (FPRs). |
| Approach: | They propose a Conformal Prediction framework that constrains the upper bound of false positive rates and introduces a real-time detection framework. |
| Outcome: | The proposed framework reduces false positive rates and improves detection performance. |
Unsupervised Dual Paraphrasing for Two-stage Semantic Parsing (2020.acl-main)
Copied to clipboard
| Challenge: | Existing semantic parsing frameworks rely on nontrivial human labor to generate canonical utterances. |
| Approach: | They propose a framework that uses an unsupervised paraphrase model to parse canonical utterances. |
| Outcome: | The proposed framework is effective and compatible with supervised training. |
AutoMedEval: Harnessing Language Models for Automatic Medical Capability Evaluation (2025.acl-long)
Copied to clipboard
Xiechi Zhang, Zetian Ouyang, Linlin Wang, Gerard De Melo, Zhu Cao, Xiaoling Wang, Ya Zhang, Yanfeng Wang, Liang He
| Challenge: | Existing evaluation methods based on large language models (LLMs) are expensive and lack expertise due to limitations in human expertise. |
| Approach: | They propose an open-source automatic evaluation model with 13B parameters specifically engineered to measure the question-answering proficiency of medical LLMs. |
| Outcome: | The proposed model surpasses baselines in terms of correlation with human judgments. |
A Chain-of-Task Framework for Instruction Tuning of LLMs Based on Chinese Grammatical Error Correction (2025.coling-main)
Copied to clipboard
| Challenge: | Existing approaches to address Grammatical Error Correction (GEC) tasks are based on large scale labeled data, which leads to extremely high data annotation costs. |
| Approach: | They propose a Chain-of-Task framework to reduce over-correction in large language models . they propose supervised fine-tuning strategy and an algorithm for automatic dataset annotation . |
| Outcome: | The proposed framework achieves state-of-the-art on both FCGEC (in-domain) and NaCGEC (out-of domain) test sets. |
Jailbreak Open-Sourced Large Language Models via Enforced Decoding (2024.acl-long)
Copied to clipboard
Hangfan Zhang, Zhimeng Guo, Huaisheng Zhu, Bochuan Cao, Lu Lin, Jinyuan Jia, Jinghui Chen, Dinghao Wu
| Challenge: | Existing studies show that Large Language Models can be misused to generate undesired content. |
| Approach: | They propose to use large language models to manipulate the generation process to generate undesired content without heavy computations or prompt designs. |
| Outcome: | The proposed method shows that open-sourced large language models could be misused to generate undesired content without heavy computations or prompt designs. |
Exploring Schema Generalizability of Text-to-SQL (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing text-to-SQL models are limited in their generalizability, despite their performance being over-estimated. |
| Approach: | They propose a framework to generate novel text-to-SQL data via automatic and synchronous (DS, SQL) pair altering. |
| Outcome: | The proposed framework generates text-to-SQL data via automatic and synchronous (DS, SQL) pair altering. |
Zero-shot Generalization in Dialog State Tracking through Generative Question Answering (2021.eacl-main)
Copied to clipboard
| Challenge: | Existing methods for Dialog State Tracking do not generalize well to new domains and unseen slots. |
| Approach: | They propose an ontology-free framework that queries for unseen constraints and slots in multi-domain task-oriented dialogs using a conditional language model pre-trained on substantive English sentences. |
| Outcome: | The proposed framework improves goal accuracy in zero-shot domain adaptation settings by up to 9% over the previous state-of-the-art on the MultiWOZ 2.1 dataset. |
The Microsoft Toolkit of Multi-Task Deep Neural Networks for Natural Language Understanding (2020.acl-demos)
Copied to clipboard
Xiaodong Liu, Yu Wang, Jianshu Ji, Hao Cheng, Xueyun Zhu, Emmanuel Awa, Pengcheng He, Weizhu Chen, Hoifung Poon, Guihong Cao, Jianfeng Gao
| Challenge: | MT-DNN is an open-source natural language understanding toolkit . it allows researchers and developers to train customized deep learning models . |
| Approach: | They present MT-DNN, an open-source natural language understanding toolkit . it is designed to facilitate rapid customization for a broad spectrum of NLU tasks . MT supports multi-task knowledge distillation, which can substantially compress a deep neural model without significant performance drop. |
| Outcome: | The proposed model can significantly compress a large model without significant performance drop. |
Hence, Socrates is mortal: A Benchmark for Natural Language Syllogistic Reasoning (2023.findings-acl)
Copied to clipboard
Yongkang Wu, Meng Han, Yutao Zhu, Lei Li, Xinyu Zhang, Ruofei Lai, Xiaoguang Li, Yuanhang Ren, Zhicheng Dou, Zhao Cao
| Challenge: | SylloBase is a benchmark for syllogistic reasoning, a critical capability widely required in natural language understanding tasks, such as text entailment and question answering. |
| Approach: | They propose to use a benchmark to learn syllogistic reasoning on a set of templates and to use them to generate and understand slogisms. |
| Outcome: | The proposed benchmark covers a complete taxonomy of syllogism reasoning patterns, and contains both automatically and manually constructed samples. |
Generating then Refining for Reliable Knowledge Base Question Answering (2026.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge base question answering methods generate LFs that are non-executable due to semantic hallucination issue of large language models. |
| Approach: | They propose a "generate-verify-refine" framework for reliable LF generation . they propose ARI-KBQA to generate query paths based on hop-by-hop reasoning . |
| Outcome: | The proposed framework significantly improves model performance with a reduced search space . ARI-KBQA can generate LFs that are non-executable due to semantic hallucination issue . |
LGESQL: Line Graph Enhanced Text-to-SQL Model with Mixed Local and Non-Local Relations (2021.acl-long)
Copied to clipboard
| Challenge: | Existing methods to encode text-to-SQL data are node-centric and ignore semantics embedded in the topological structure of edges. |
| Approach: | They propose a Line Graph Enhanced Text-to-SQL model to mine relational features without constructing meta-paths. |
| Outcome: | The proposed model achieves state-of-the-art on the cross-domain text-to-SQL benchmark Spider at the time of writing. |
InquireMobile: Teaching VLM-based Mobile Agent to Request Human Assistance via Reinforcement Fine-Tuning (2026.acl-long)
Copied to clipboard
Qihang Ai, Pi Bu, Yue Cao, Yingyao Wang, Jihao Gu, Jingxuan Xing, Zekun Zhu, Wei Jiang, Zhicheng Zheng, Jun Song, Yuning Jiang
| Challenge: | Recent advances in Vision-Language Models (VLMs) have enabled mobile agents to perceive and interact with real-world mobile environments based on human instructions. |
| Approach: | They propose a vision-language model that actively seeks human confirmation at critical decision points and a model inspired by reinforcement learning. |
| Outcome: | The proposed model achieves an improvement of 46.8% in inquiry success rate and the best overall success rate among existing baselines on InquireBench. |
Multi-Faceted Self-Consistent Preference Alignment for Query Rewriting in Conversational Search (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to rewrite ambiguous queries ignore feedback from query rewriting, passage retrieval and response generation in the rewritten process. |
| Approach: | They propose to construct self-consistent preference alignment data to generate more diverse rewritten queries. |
| Outcome: | The proposed method is effective in both in- and out-of-distribution scenarios. |
Sparse Activation Editing for Reliable Instruction Following in Narratives (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing benchmarks fail to capture the challenges of instruction following in complex narrative contexts. |
| Approach: | They propose a training-free framework that identifies and edits instruction-relevant neurons using only natural language instructions without requiring labelled data. |
| Outcome: | The proposed framework improves instruction following by identifying and editing instruction-relevant neurons using only natural language instructions, without requiring labelled data. |
RankPrompt: Step-by-Step Comparisons Make Language Models Better Reasoners (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing solutions to reasoning tasks require extensive human annotations or fail in scenarios with inconsistent responses. |
| Approach: | They propose a new method that enables LLMs to self-rank their responses without additional resources. |
| Outcome: | The proposed method improves reasoning performance of ChatGPT and GPT-4 with 13% improvement over existing methods. |
Exons-Detect: Identifying and Amplifying Exonic Tokens via Hidden-State Discrepancy for Robust AI-Generated Text Detection (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for AI-generated text detection assume uniform token contributions, making them less robust under short sequences or localized token modifications. |
| Approach: | They propose a training-free method for AI-generated text detection based on an exon-aware token reweighting perspective. |
| Outcome: | The proposed method achieves state-of-the-art detection performance and robustness to adversarial attacks and varying input lengths. |
TokenPenalty: Alleviating Attention Sinks and Positional Decay in LVLMs (2026.findings-acl)
Copied to clipboard
| Challenge: | Multimodal large language models (MLLMs) often hallucinate due to two relevant phenomena: massive activation phenomenon and positional information decay. |
| Approach: | They propose a token-level intervention strategy that dynamically suppresses irrelevant visual tokens while preserving key contextual signals. |
| Outcome: | Experiments show that TokenTruth significantly improves factual consistency across MLLMs on standard image understanding benchmarks. |
ShadowGNN: Graph Projection Neural Network for Text-to-SQL Parser (2021.naacl-main)
Copied to clipboard
| Challenge: | Existing semantic parsing models struggle to adapt to unseen database schemas . a new architecture, ShadowGNN, processes schemas at abstract and semantic levels . |
| Approach: | They propose a new architecture which processes schemas at abstract and semantic levels. |
| Outcome: | The proposed architecture outperforms state-of-the-art models on a text-to-sql benchmark . it uses domain-independent representations to extract logical linking between question and schema . |
PGPO: Enhancing Agent Reasoning via Pseudocode-style Planning Guided Preference Optimization (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing LLM agents generate verbose and inefficient natural language plans to guide reasoning, which restricts agents’ ability to generalize across similar tasks. |
| Approach: | They propose a pseudocode-style planning guide optimization method that captures the structural logic of reasoning and uses two planning-oriented rewards to enhance agent learning. |
| Outcome: | The proposed method outperforms existing LLM agents on representative agent benchmarks and outperformed the current leading baselines. |
InsightEval: An Expert-Curated Benchmark for Assessing Insight Discovery in LLM-Driven Data Agents (2026.findings-acl)
Copied to clipboard
Zhenghao Zhu, Yuanfeng Song, Xing Chen, Chengzhong Liu, Cui Yakun, Caleb Chen Cao, Sirui Han, Yike Guo
| Challenge: | Existing frameworks for data analysis and insight exploration are lacking in terms of benchmarks . existing frameworks suffer from format inconsistencies, poorly conceived objectives, and redundant insights. |
| Approach: | They propose a data-curation pipeline to construct a new dataset named InsightEval. |
| Outcome: | The proposed benchmarks highlight prevailing challenges in automated insight discovery and raise key findings to guide future research. |
APPSI-139: A Parallel Corpus of English Application Privacy Policy Summarization and Interpretation (2026.acl-long)
Copied to clipboard
Pengyun Zhu, Qiheng Sun, Long Wen, Yanbo Wang, Yang Cao, Junxu Liu, Deyi Xiong, Jinfei Liu, Zhibo Wang, Kui Ren
| Challenge: | a lack of high-quality English privacy policy corpus optimized for legal clarity and readability is limiting translation of privacy policies . 139 privacy policies are often considered "incomprehensible" due to technical jargon, legal language, and convoluted grammatical structures. |
| Approach: | They propose a high-quality English privacy policy corpus annotated by domain experts . they propose APPSI-139 to summarize and interpret privacy policies in English . |
| Outcome: | The proposed framework outperforms large language models in terms of readability and accuracy. |
Discourse Coherence and Response-Guided Context Rewriting for Multi-Party Dialogue Generation (2026.acl-long)
Copied to clipboard
| Challenge: | Existing research on multi-party dialogue generation has focused on structural information inherent in dialogues, but colloquial expressions and incomplete utterances often impede comprehension and weaken the fidelity of dialogue structure representations. |
| Approach: | They propose a framework to improve multi-party dialogue generation through dialogue context rewriting using two complementary feedback signals to construct preference data for both context & response generation. |
| Outcome: | The proposed framework improves multi-party dialogue generation through dialogue context rewriting. |
Incomplete Utterance Rewriting with Editing Operation Guidance and Utterance Augmentation (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing generation methods on Incomplete Utterance Rewriting (IUR) can generate coherent utterances, but they often include irrelevant and redundant tokens in rewritten utteras . |
| Approach: | They propose a multi-task learning framework that uses editing operation labels to guide generation model to focus on critical tokens in dialogue context. |
| Outcome: | The proposed model outperforms state-of-the-art models on open-domain and task-oriented dialogues on three datasets. |
SPM: A Split-Parsing Method for Joint Multi-Intent Detection and Slot Filling (2023.acl-industry)
Copied to clipboard
| Challenge: | Existing studies focus on utterances with a single intent, but lack the ability to assign slots to each corresponding intent. |
| Approach: | They propose a split-parsing method for joint intent detection and slot filling . they split an input sentence into multiple sub-sentences which contain a single-intent . |
| Outcome: | The proposed method improves on three multi-intent datasets on multi-tasks. |
Revealing the Parallel Multilingual Learning within Large Language Models (2024.emnlp-main)
Copied to clipboard
Yongyu Mu, Peinan Feng, Zhiquan Cao, Yuzhang Wu, Bei Li, Chenglong Wang, Tong Xiao, Kai Song, Tongran Liu, Chunliang Zhang, JingBo Zhu
| Challenge: | Large language models (LLMs) can handle multilingual and cross-lingual text within a single input; however, previous studies focusing on using English as the pivot language to enhance language understanding and reasoning focus on using multiple languages. |
| Approach: | They propose to use parallel multilingual input to enhance the model's comprehension of the input and to examine how multilingual processing affects prediction. |
| Outcome: | The proposed model can handle multilingual and cross-lingual text within a single input, but previous studies focused on using English as the pivot language to enhance language understanding and reasoning. |
Same Claim, Different Judgment: Benchmarking Scenario-Induced Bias in Multilingual Financial Misinformation Detection (2026.findings-acl)
Copied to clipboard
Zhiwei Liu, Yupeng Cao, Yuechen Jiang, Mohsinul Kabir, Polydoros Giannouris, Chen Xu, Ziyang Xu, Tianlei Zhu, Md. Tariquzzaman, Triantafillos Papadopoulos, Yan Wang, Lingfei Qian, Xueqing Peng, Zhuohan Xie, Ye Yuan, Saeed Almheiri, Abdulrazzaq Alnajjar, Ming-Bin Chen, Harry Stuart, Paul Thompson, Prayag Tiwari, Alejandro Lopez-Lira, Xue Liu, Jimin Huang, Sophia Ananiadou
| Challenge: | Existing research on LLM biases has focused on direct questioning or general-purpose settings . pronounced behavioral biase despite their growing deployment in financial analysis, forecasting, and decision support. |
| Approach: | They propose a benchmark to evaluate behavioral biases of large language models in MFMD . they use a multilingual financial misinformation dataset to integrate these with misinformation claims . |
| Outcome: | The proposed benchmark evaluates behavioral biases of large language models across economic scenarios. |
See the World, Discover Knowledge: A Chinese Factuality Evaluation for Large Vision Language Models (2025.findings-acl)
Copied to clipboard
Jihao Gu, Yingyao Wang, Pi Bu, Chen Wang, Ziming Wang, Tengtao Song, Donglai Wei, Jiale Yuan, Yingxiu Zhao, Yancheng He, Shilong Li, Jiaheng Liu, Meng Cao, Jun Song, Yingshui Tan, Xiang Li, Wenbo Su, Xiaoyong Zhu, Bo Zheng
| Challenge: | Existing models for large vision language models do not fully reflect their knowledge capacity and reliability, resulting in erroneous outputs that do not align with the image content or provide answers lacking knowledge evidence. |
| Approach: | They propose a Chinese-based benchmark for visual factuality across 8 major topics and 56 subtopics and a multi-hop question construction. |
| Outcome: | The proposed model decouples visual factuality into two parts: seeing the world and discovering knowledge. |
Line Graph Enhanced AMR-to-Text Generation with Mix-Order Graph Attention Networks (2020.acl-main)
Copied to clipboard
| Challenge: | Existing graph-to-sequence approaches use graph neural networks as encoders, but they lack the structure information needed to translate AMR into the graph-based data. |
| Approach: | They propose a graph-to-sequence task which aims to recover natural language from Abstract Meaning Representations (AMR) they adopt graph attention networks with higher-order neighborhood information to explore the edge relations in AMR graphs. |
| Outcome: | The proposed framework achieves state-of-the-art performance on English AMR benchmark datasets and is able to translate the AMR semantics into the natural language. |
I²B-LPO: Latent Policy Optimization via Iterative Information Bottleneck (2026.acl-long)
Copied to clipboard
Huilin Deng, Hongchen Luo, Yue Zhu, Long Li, Zhuoyue Chen, Xinghao Zhao, Ming LI, Chuyang Zhao, Jihai Zhang, MengChang Wang, Yang Cao, Yu Kang
| Challenge: | Existing methods for large language model reasoning suffer from exploration collapse due to the semantic homogeneity of random rollouts. |
| Approach: | They propose to use latent policy optimization via iterative information bottleneck to optimize reasoning trajectories by diversifying reasoning . |
| Outcome: | Empirical results show that the proposed method achieves state-of-the-art performance with margins of up to 5.3% in accuracy and 7.4% in diversity metrics. |
Across Programming Language Silos: A Study on Cross-Lingual Retrieval-Augmented Code Generation (2026.findings-acl)
Copied to clipboard
Qiming Zhu, Jialun Cao, Xuanang Chen, Weili Zhang, Yaojie Lu, Hongyu Lin, Xianpei Han, Le Sun, Shing-Chi Cheung
| Challenge: | Current research on large language models with retrieval-augmented code generation (RACG) has focused on single-language settings, leaving their cross-lingual effectiveness underexplored. |
| Approach: | They construct a dataset covering 13 PLs with nearly 14K instances to study cross-lingual code knowledge transfer in RACG. |
| Outcome: | The proposed model shows unequal cross-lingual knowledge transfer even with direct injection and shows limited reliance on natural language information embedded in code when equipped with a code-specific retriever. |
INVESTORBENCH: A Benchmark for Financial Decision-Making Tasks with LLM-based Agent (2025.acl-long)
Copied to clipboard
Haohang Li, Yupeng Cao, Yangyang Yu, Shashidhar Reddy Javaji, Zhiyang Deng, Yueru He, Yuechen Jiang, Zining Zhu, K.p. Subbalakshmi, Jimin Huang, Lingfei Qian, Xueqing Peng, Jordan W. Suchow, Qianqian Xie
| Challenge: | Recent advances have underscored the potential of large language model (LLM)-based agents in financial decision-making. |
| Approach: | They propose to evaluate LLM agents using 13 different LLMs as backbone models across various market environments and tasks. |
| Outcome: | The proposed framework assesses the reasoning and decision-making capabilities of 13 different LLMs across various market environments and tasks. |
ML-LMCL: Mutual Learning and Large-Margin Contrastive Learning for Improving ASR Robustness in Spoken Language Understanding (2023.findings-acl)
Copied to clipboard
| Challenge: | Despite efforts to improve ASR robustness, errors from pipeline approaches can lead to error propagation. |
| Approach: | They propose a framework for improving ASR robustness in SLU by using mutual learning and large-margin contrastive learning. |
| Outcome: | The proposed framework outperforms existing models and achieves new state-of-the-art performance on three datasets. |
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. |
Semantic Parsing with Dual Learning (P19-1)
Copied to clipboard
| Challenge: | Existing approaches to parse natural language queries are limited by lack of labeled data and constrained decoding. |
| Approach: | They propose a semantic parsing framework with the dual learning algorithm that makes full use of data through a dual-learning game. |
| Outcome: | The proposed approach achieves state-of-the-art performance on ATIS dataset and gets competitive performance on overnight dataset. |
RSVP: Reasoning Segmentation via Visual Prompting and Multi-modal Chain-of-Thought (2025.acl-long)
Copied to clipboard
Yi Lu, Jiawang Cao, Yongliang Wu, Bozheng Li, Licheng Tang, Yangguang Ji, Chong Wu, Jay Wu, Wenbo Zhu
| Challenge: | Recent advances in multi-modal learning have enhanced MLLMs' ability to reason about visual content. |
| Approach: | They propose a framework that unifies multi-step multimodal reasoning with grounded visual understanding. |
| Outcome: | The proposed framework surpasses state-of-the-art methods by +6.5 gIoU and +9.2 cIou on ReasonSeg and achieves 49.7 mAP on SegInW under zero-shot settings. |
Enhancing Chinese Pre-trained Language Model via Heterogeneous Linguistics Graph (2022.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that pre-trained Chinese language models ignore linguistics knowledge to learn representations. |
| Approach: | They propose a task-free enhancement module to integrate linguistics knowledge into Chinese pre-trained language models. |
| Outcome: | The proposed model improves Chinese pre-trained language models on 6 tasks with 10 benchmark datasets. |
Zoom Out and Observe: News Environment Perception for Fake News Detection (2022.acl-long)
Copied to clipboard
| Challenge: | Existing methods for fake news detection "zoom in" to verify content with knowledge sources or check readers’ replies to posts but neglect information in the external news environment where a fake news post is created and disseminated. |
| Approach: | They propose a framework to capture news environment signals and a module to perceive useful signals and assist final prediction. |
| Outcome: | The proposed framework can improve the performance of basic fake news detectors by capturing the environmental signals of news posts and analyzing the results. |
Learn over Past, Evolve for Future: Forecasting Temporal Trends for Fake News Detection (2023.acl-industry)
Copied to clipboard
| Challenge: | Existing work on fake news detection does not consider the temporal shift issue caused by the rapidly-evolving nature of news data. |
| Approach: | They propose a framework to forecast temporal patterns of news data and guide detector to fast adapt to future distributions. |
| Outcome: | The proposed framework forecasts temporal distribution patterns and guides detector to fast adapt to future distribution. |
AdaReTaKe: Adaptive Redundancy Reduction to Perceive Longer for Video-language Understanding (2025.findings-acl)
Copied to clipboard
| Challenge: | Multimodal Large Language Models (MLLMs) are limited by context length when processing long videos. |
| Approach: | They propose a training-free method that flexibly reduces redundancy by allocating compression ratios among time and model layers with theoretical guarantees. |
| Outcome: | Experiments on videoMME, MLVU, LongVideoBench, and LVBench show that AdaRETAKE outperforms existing methods by 2.3% and 2.8% for 7B and 72B models. |
Scaling Down, Serving Fast: Compressing and Deploying Efficient LLMs for Recommendation Systems (2025.emnlp-industry)
Copied to clipboard
Kayhan Behdin, Ata Fatahibaarzi, Qingquan Song, Yun Dai, Aman Gupta, Zhipeng Wang, Hejian Sang, Shao Tang, Gregory Dexter, Sirou Zhu, Siyu Zhu, Tejas Dharamsi, Vignesh Kothapalli, Zhoutong Fu, Yihan Cao, Pin-Lun Hsu, Fedor Borisyuk, Natesh S. Pillai, Luke Simon, Rahul Mazumder
| Challenge: | Large language models (LLMs) have demonstrated remarkable performance across a wide range of industrial applications. |
| Approach: | They propose two techniques for training and deploying small language models that deliver high performance for a variety of industry use cases. |
| Outcome: | The proposed techniques retain much of the quality of larger models while reducing training/serving costs and latency. |
Improving Autoregressive Grammatical Error Correction with Non-autoregressive Models (2023.findings-acl)
Copied to clipboard
| Challenge: | Autoregressive models assign low probabilities to tokens that need corrections . grammatical error correction (GEC) is widely applied to natural language processing tasks . |
| Approach: | They propose to use a non-autoregressive model as an auxiliary model to train GEC models to correct grammatical errors in sentences. |
| Outcome: | The proposed method outperforms baselines on English and Chinese GEC tasks significantly. |
ICR: Iterative Clarification and Rewriting for Conversational Search (2025.emnlp-main)
Copied to clipboard
| Challenge: | Conversational Query Rewriting (CQR) is a key step in conversational question answering . it aims to rewrite vague queries into de-contextualized queries, thereby promoting conversational search. |
| Approach: | They propose an iterative rewriting scheme that pivots on clarification questions . they propose to rewrite queries into de-contextualized queries to promote conversational search . |
| Outcome: | The proposed framework improves retrieval performance on two popular datasets. |
DocEE: A Large-Scale and Fine-grained Benchmark for Document-level Event Extraction (2022.naacl-main)
Copied to clipboard
| Challenge: | Existing datasets focus on sentence-level event extraction, but document-level EE is limited due to the lack of large-scale and practical training and evaluation datasets. |
| Approach: | They propose a document-level event extraction dataset with 27,000+ events and 180,000+ arguments. |
| Outcome: | The proposed dataset includes 27,000+ events, 180,000+ arguments and large-scale manual annotations, fine-grained argument types and application-oriented settings. |
Coarse-to-Fine: Hierarchical Multi-task Learning for Natural Language Understanding (2022.coling-1)
Copied to clipboard
Zhaoye Fei, Yu Tian, Yongkang Wu, Xinyu Zhang, Yutao Zhu, Zheng Liu, Jiawen Wu, Dejiang Kong, Ruofei Lai, Zhao Cao, Zhicheng Dou, Xipeng Qiu
| Challenge: | Existing methods to learn downstream tasks by stitches skill block lack rationality and interpretation. |
| Approach: | They propose a hierarchical framework with a coarse-to-fine paradigm for generalized text representations from the large-scale corpus. |
| Outcome: | The proposed model learns basic language properties from all tasks and boosts performance on relevant tasks. |
Enabling Agents to Communicate Entirely in Latent Space (2026.acl-long)
Copied to clipboard
Zhuoyun Du, Runze Wang, Huiyu Bai, Zouying Cao, Xiaoyong Zhu, Yu Cheng, Bo Zheng, Wei Chen, Haochao Ying
| Challenge: | Natural language is the de facto communication medium for LLM-based agents, but it presents a fundamental constraint . natural language downsampling limits the depth and nuance of information that can be transmitted . et al.: inter-agent latent space communication is a promising paradigm for solving complex tasks . |
| Approach: | They propose a paradigm that leverages the last hidden states of an LLM as a representation of its thought for direct communication. |
| Outcome: | The proposed paradigm outperforms fine-tuned chain-of-thought prompting and single-agent baselines even across heterogeneous models. |
Augmenting Large Language Model Translators via Translation Memories (2023.findings-acl)
Copied to clipboard
Yongyu Mu, Abudurexiti Reheman, Zhiquan Cao, Yuchun Fan, Bei Li, Yinqiao Li, Tong Xiao, Chunliang Zhang, Jingbo Zhu
| Challenge: | Using translation memories (TMs) as prompts is a promising approach to in-context learning of machine translation models. |
| Approach: | They propose to use translation memories (TMs) as prompts to prompt large language models (LLMs) they find that the ability of LLMs to "understand" prompts is helpful . |
| Outcome: | The results are comparable to state-of-the-art NMT systems with bilingual data and are tuned on downstream tasks. |
ACE-M3: Automatic Capability Evaluator for Multimodal Medical Models (2025.coling-main)
Copied to clipboard
| Challenge: | Existing metrics for multimodal large language models only focus on token overlap and may not align with human judgment. |
| Approach: | They propose an open-source model that assesses the question answering abilities of multimodal large language models. |
| Outcome: | Experiments show that the ACE-M3 model performs better than existing models and is more reliable than existing metrics. |
MolSafeEval: A Benchmark for Uncovering Safety Risks in AI-Generated Molecules (2026.findings-acl)
Copied to clipboard
| Challenge: | generative models may produce molecules with toxic, reactive, or otherwise hazardous characteristics. |
| Approach: | They propose a benchmark to evaluate and analyze the safety risks of molecular generation. |
| Outcome: | The proposed benchmark aims to evaluate and analyze the safety risks of molecular generation. |
SafeLawBench: Towards Safe Alignment of Large Language Models (2025.findings-acl)
Copied to clipboard
Chuxue Cao, Han Zhu, Jiaming Ji, Qichao Sun, Zhenghao Zhu, Wu Yinyu, Josef Dai, Yaodong Yang, Sirui Han, Yike Guo
| Challenge: | Recent studies indicate that large language models (LLMs) may exhibit risks, including threats to the protection of private data and the generation of hallucinations. |
| Approach: | They propose to evaluate LLMs from a legal perspective using the SafeLawBench benchmark. |
| Outcome: | The proposed framework categorizes safety risks into three levels based on legal standards and includes 24,860 multi-choice questions and 1,106 open-domain question-answering tasks. |
Few-shot Temporal Pruning Accelerates Diffusion Models for Text Generation (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing acceleration methods for text generation ignore the importance of the distribution of sampling steps, resulting in slow sampling rates. |
| Approach: | They propose a technique to accelerate diffusion models for text generation without additional training by using a Bayesian optimization approach. |
| Outcome: | The proposed technique achieves 400x acceleration even with minimal sampling steps after down to less than 1 minute of optimization yielding a competitive performance even with minimum sampling steps. |
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. |
Teaching Language Models to Self-Improve by Learning from Language Feedback (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) generate content that can be untruthful or harmful. |
| Approach: | They propose a method that leverages model feedback for alignment . they use a base language model to generate initial responses, critiqued and refined . |
| Outcome: | The proposed method outperforms strong baselines across diverse tasks and model sizes. |
HeteroSpec: Leveraging Contextual Heterogeneity for Efficient Speculative Decoding (2026.acl-long)
Copied to clipboard
| Challenge: | Autoregressive decoding limits the inference throughput of Large Language Models due to its sequential dependency. |
| Approach: | They propose a framework that allocates verification effort in proportion to candidate uncertainty. |
| Outcome: | Speculative decoding achieves an average speedup over state-of-the-art methods . a small subset of high-confidence predictions accounts for most successful verifications . |
FinEval: A Chinese Financial Domain Knowledge Evaluation Benchmark for Large Language Models (2025.naacl-long)
Copied to clipboard
Xin Guo, Haotian Xia, Zhaowei Liu, Hanyang Cao, Zhi Yang, Zhiqiang Liu, Sizhe Wang, Jinyi Niu, Chuqi Wang, Yanhui Wang, Xiaolong Liang, Xiaoming Huang, Bing Zhu, Zhongyu Wei, Yun Chen, Weining Shen, Liwen Zhang
| Challenge: | Large language models have demonstrated outstanding performance in various natural language processing tasks, but their security capabilities in the financial domain have not been explored. |
| Approach: | They propose to use a benchmark to evaluate large language models' financial domain knowledge and practical abilities. |
| Outcome: | The proposed benchmark evaluates large language models' financial domain knowledge and practical abilities. |
CoUDA: Coherence Evaluation via Unified Data Augmentation (2024.naacl-long)
Copied to clipboard
| Challenge: | Existing data augmentations for coherence evaluation rely on heuristic rules and lack designing criteria. |
| Approach: | They propose a data augmentation framework that breaks down coherence into global and local aspects and designs augmentation strategies for both aspects. |
| Outcome: | The proposed framework surpasses recent models in scoring and ranking tasks with 233M parameters. |
Conjoin after Decompose: Improving Few-Shot Performance of Named Entity Recognition (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing prompt-based NER models fail to detect entity boundaries, causing performance degradation. |
| Approach: | They propose a model which consists of a BART encoder and a parabiotic decoder and propose ' boundary expansion strategy' to enhance the model's capability in entity type classification. |
| Outcome: | The proposed model can achieve significant performance gains over state-of-the-art models. |
Dr3: Ask Large Language Models Not to Give Off-Topic Answers in Open Domain Multi-Hop Question Answering (2024.lrec-main)
Copied to clipboard
Yuan Gao, Yiheng Zhu, Yuanbin Cao, Yinzhi Zhou, Zhen Wu, Yujie Chen, Shenglan Wu, Haoyuan Hu, Xinyu Dai
| Challenge: | Open Domain Multi-Hop Question Answering (ODMHQA) is one of the most challenging tasks in Natural Language Processing (NLP) |
| Approach: | They propose a mechanism that leverages the intrinsic capabilities of Large Language Models to judge whether the generated answers are off-topic. |
| Outcome: | The proposed method reduces the occurrence of off-topic answers by nearly 13%, improving the performance in Exact Match (EM) by nearly 3% compared to the baseline method without the Dr3 mechanism. |
Enhancing Reinforcement Learning with Dense Rewards from Language Model Critic (2024.emnlp-main)
Copied to clipboard
| Challenge: | Reinforcement learning (RL) can align language models with non-differentiable reward signals, such as human preferences, but the sparsity of these signals can lead to inefficient and unstable learning. |
| Approach: | They propose a framework that utilizes the critique capability of Large Language Models to produce intermediate-step rewards during RL training. |
| Outcome: | The proposed framework improves sample efficiency and the overall performance of the policy model, supported by both automatic and human evaluation. |
Improving Fake News Detection of Influential Domain via Domain- and Instance-Level Transfer (2022.coling-1)
Copied to clipboard
| Challenge: | Social media spreads both real news and fake news in various domains including politics, health, entertainment, etc. |
| Approach: | They propose a Domain- and Instance-level Transfer Framework for Fake News Detection which could improve the performance of specific target domains. |
| Outcome: | The proposed framework improves performance of target domains by hurting other domains, resulting in unsatisfactory performance in the target domain. |
Transferable and Efficient: Unifying Dynamic Multi-Domain Product Categorization (2023.acl-industry)
Copied to clipboard
| Challenge: | e-commerce platforms are encountering increasingly complex product categorization scenarios . multiple business domains correspond to different category taxonomies, with different depths and distinct literal expressions of category names. |
| Approach: | They propose a taxonomy-agnostic framework that calculates semantic relatedness between product titles and category names in the vector space. |
| Outcome: | The proposed framework outperforms strong baselineson three dynamic multi-domain product categorization tasks. |
Two-stage Incomplete Utterance Rewriting on Editing Operation (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods to generate rewritten utterances based on dialogue context ignore coreference and ellipsis in dialogues. |
| Approach: | They propose a framework where the first stage generates editing operations and the second stage rewrites incomplete utterances utilizing the generated editing operations. |
| Outcome: | The proposed framework outperforms the existing models on three IUR datasets. |
BeautifulPrompt: Towards Automatic Prompt Engineering for Text-to-Image Synthesis (2023.emnlp-industry)
Copied to clipboard
| Challenge: | Recent text-to-image models require multiple passes of prompt engineering by humans to produce satisfactory results for real-world applications. |
| Approach: | They propose a deep generative model to generate high-quality prompts from raw descriptions using visual feedback. |
| Outcome: | The proposed model produces high-quality prompts from simple raw descriptions . it can be integrated to a cloud-native AI platform to provide better image generation service in the cloud. |
Mobile-R1: Towards Interactive Capability for VLM-Based Mobile Agent via Systematic Training (2026.acl-long)
Copied to clipboard
Jihao Gu, Qihang Ai, Yingyao Wang, Pi Bu, Jingxuan Xing, Yue Cao, Zekun Zhu, Wei Jiang, Ziming Wang, Yingxiu Zhao, Ming-Liang Zhang, Jun Song, Yuning Jiang, Bo Zheng
| Challenge: | Existing approaches to training agents for visual-language models trap them in local optima, hindering exploration and error correction with the environment. |
| Approach: | They propose a hierarchical training recipe that bridges atomic action execution and strategic task completion. |
| Outcome: | The proposed training recipe bridges atomic action execution and strategic task completion. |