THOR-MoE: Hierarchical Task-Guided and Context-Responsive Routing for Neural Machine Translation
机构 * Pattern Recognition Center, WeChat AI, Tencent Inc（模式识别中心、微信AI、腾讯公司）
Comments Accepted to ACL 2025 main conference
期刊&会议
ACL
Annual Meeting of the Association for Computational Linguistics · 会议 · Natural Language Processing
Online Iterative Self-Alignment for Radiology Report Generation
机构 * East China University of Science and Technology（东华大学） ； Tsinghua University（清华大学） ； Institute of Artificial Intelligence (TeleAI)（人工智能研究院）
Comments Accepted by ACL 2025 Main
MathAgent: Leveraging a Mixture-of-Math-Agent Framework for Real-World Multimodal Mathematical Error Detection
机构 * Squirrel Ai Learning ； The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州）） ； The Hong Kong University of Science and Technology（香港科学与技术大学） ； University of Illinois at Chicago（伊利诺伊大学香槟分校）
Comments Accepted by The 63rd Annual Meeting of the Association for Computational Linguistics (ACL Industry 2025, Oral Presentation)
SQLong: Enhanced NL2SQL for Longer Contexts with LLMs
机构 * Oracle Corporation（Oracle公司）
Comments Accepted to Table Representation Learning Workshop at ACL 2025
Who Taught You That? Tracing Teachers in Model Distillation
机构 * Northeastern University（东北大学）
Comments Findings of ACL 2025
LongDPO: Unlock Better Long-form Generation Abilities for LLMs via Critique-augmented Stepwise Information
机构 * State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University（多媒体信息处理国家重点实验室，计算机科学学院，北京大学） ； Pattern Recongnition Center, WechatAI, Tencent Inc（图案识别中心，微信AI，腾讯公司） ； Dept. of Comp. Sci. & Tech., Tsinghua University（计算机科学与技术系，清华大学）
Comments ACL 2025
InternLM-XComposer2.5-Reward: A Simple Yet Effective Multi-Modal Reward Model
机构 * Shanghai Artificial Intelligence Laboratory（上海人工智能实验室） ； The Chinese University of Hong Kong（香港中文大学） ； Shanghai Jiao Tong University（上海交通大学） ； Nanjing University（南京大学） ； Fudan University（复旦大学） ； Nanyang Technological University（南洋理工大学）
Comments ACL 2025 Findings
ToolHop: A Query-Driven Benchmark for Evaluating Large Language Models in Multi-Hop Tool Use
机构 * Fudan University（复旦大学） ； ByteDance（字节跳动） ； Shanghai Collaborative Innovation Center of Intelligent Visual Computing（上海智能视觉计算协同创新中心）
Comments Accepted by ACL 2025 Main Conference
A Survey of Mathematical Reasoning in the Era of Multimodal Large Language Model: Benchmark, Method & Challenges
Comments Accepted by The 63rd Annual Meeting of the Association for Computational Linguistics (ACL Findings 2025)
Velocitune: A Velocity-based Dynamic Domain Reweighting Method for Continual Pre-training
机构 * The University of Manchester（曼彻斯特大学） ； Microsoft（微软） ； Tsinghua University（清华大学）
Comments Accepted to ACL 2025
Texts or Images? A Fine-grained Analysis on the Effectiveness of Input Representations and Models for Table Question Answering
机构 * Bosch Center for Artificial Intelligence（博世人工智能中心） ； Hochschule der Medien（媒体高等学校） ； University of Augsburg（艾格斯堡大学）
Comments Accepted at ACL25 (Findings)
Dolphin: Document Image Parsing via Heterogeneous Anchor Prompting
机构 * ByteDance（字节跳动）
Comments Accepted to ACL 2025
Improve Language Model and Brain Alignment via Associative Memory
机构 * College of Artificial Intelligence, Nanjing University of Aeronautics and Astronautics（人工智能学院，南京航空航天大学） ； The Key Laboratory of Brain-Machine Intelligence Technology, Ministry of Education（脑机智能技术重点实验室，教育部）
Comments Accepted by Findings of ACL 2025
Cross-Lingual Representation Alignment Through Contrastive Image-Caption Tuning
机构 * Department of Computer Science, George Mason University（计算机科学系，乔治·马歇尔大学）
Comments Accepted to ACL 2025 Main Conference
Combining the Best of Both Worlds: A Method for Hybrid NMT and LLM Translation
机构 * Huawei Translation Service Center（华为翻译服务中心）
Comments 9 pages, 2 figures, 9 tables, ACL 2025
Predicting Turn-Taking and Backchannel in Human-Machine Conversations Using Linguistic, Acoustic, and Visual Signals
机构 * School of Informatics, Xiamen University（厦门大学信息学院）
Comments Accepected by ACL 2025
LogicQA: Logical Anomaly Detection with Vision Language Model Generated Questions
机构 * Department of Industrial Engineering, Yonsei University（工业工程系，延世大学）
Comments Accepted Industry Track at ACL 2025
Enhancing Conversational Agents with Theory of Mind: Aligning Beliefs, Desires, and Intentions for Human-Like Interaction
机构 * UNSW Sydney（新南威尔士大学悉尼分校）
Comments Accepted to Findings of ACL 2025
TreeCut: A Synthetic Unanswerable Math Word Problem Dataset for LLM Hallucination Evaluation
机构 * Columbia University（哥伦比亚大学）
Comments Accepted to ACL 2025 Main Conference
EssayJudge: A Multi-Granular Benchmark for Assessing Automated Essay Scoring Capabilities of Multimodal Large Language Models
机构 * The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）） ； Beijing Future Brain Education Technology Co., Ltd.（北京未来脑教育科技有限公司） ； The Hong Kong University of Science and Technology（香港科技大学） ； Tsinghua University（清华大学） ； Guangxi Zhuang Autonomous Region Big Data Research Institute（广西壮族自治区大数据研究院）
Comments Accepted by ACL Findings 2025
People who frequently use ChatGPT for writing tasks are accurate and robust detectors of AI-generated text
机构 * University of Maryland, College Park（马里兰大学 College Park 分校） ； Microsoft（微软公司） ； UMass Amherst（马萨诸塞大学阿姆赫斯特分校）
Comments ACL 2025 33 pages
STATE ToxiCN: A Benchmark for Span-level Target-Aware Toxicity Extraction in Chinese Hate Speech Detection
机构 * School of Computer Science and Technology, Dalian University of Technology（计算机科学与技术学院，大连理工大学）
Comments Our paper has been accepted by ACL 2025 Findings
Cross-model Transferability among Large Language Models on the Platonic Representations of Concepts
机构 * Sichuan University, China（四川大学） ； Engineering Research Center of Machine Learning and Industry Intelligence, Ministry of Education, China（教育部机器学习与工业智能工程研究中心）
Comments ACL 2025 Main Camera Ready
Extending Complex Logical Queries on Uncertain Knowledge Graphs
机构 * Department of Mathematical Sciences, Tsinghua University, Beijing, China（清华大学数学科学系） ； CSE, HKUST, HKSAR, China（香港科技大学计算机科学与工程系） ； Department of Computer Science, Princeton University, Princeton, United States（普林斯顿大学计算机科学系）
Comments Accepted by the main conference of ACL 2025
SeedBench: A Multi-task Benchmark for Evaluating Large Language Models in Seed Science
机构 * Shanghai Artificial Intelligence Laboratory（上海人工智能实验室） ； Yazhouwan National Laboratory（Yazhouwan国家实验室） ； Shanghai Innovation Institute（上海创新研究院）
Comments Accepted by ACL 2025
An Empirical Study of Many-to-Many Summarization with Large Language Models
机构 * Pattern Recognition Center, WeChat AI, Tencent Inc, China（腾讯人工智能研究中心） ； Fudan Unversity（复旦大学） ； Waseda University（早稻田大学） ； Beijing Jiaotong University（北京交通大学） ； Zhejiang University（浙江大学）
Comments Accepted to ACL 2025 main conference
Re-identification of De-identified Documents with Autoregressive Infilling
机构 * University of Oslo Language Technology Group（奥斯陆大学语言技术组） ； Norwegian Computing Center (NR)（挪威计算中心（NR））
Comments To be presented a ACL 2025, Main, Long paper
Contrastive Prompting Enhances Sentence Embeddings in LLMs through Inference-Time Steering
机构 * State Key Laboratory for Novel Software Technology, Nanjing University, China（南京大学新型软件技术国家重点实验室）
Comments ACL 2025
Rethinking Reward Model Evaluation Through the Lens of Reward Overoptimization
Comments Accepted to ACL 2025
"Yes, My LoRD." Guiding Language Model Extraction with Locality Reinforced Distillation
机构 * The Hong Kong Polytechnic University（香港理工大学） ； The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）） ； Hong Kong Baptist University（香港 Baptist大学）
Comments To appear at ACL 25 main conference
