Dynamic Expert Specialization: Towards Catastrophic Forgetting-Free Multi-Domain MoE Adaptation
机构 * The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））
Comments EMNLP 2025 Main Conference
期刊&会议
EMNLP
Conference on Empirical Methods in Natural Language Processing · 会议 · Natural Language Processing
Benchmarking and Mitigating MCQA Selection Bias of Large Vision-Language Models
Comments Accepted to EMNLP 2025 (Main Conference)
MoRoVoc: A Large Dataset for Geographical Variation Identification of the Spoken Romanian Language
机构 * National University of Science and Technology POLITEHNICA Bucharest（波兰亚技术大学布加勒斯特分校） ； Paris 1 Panthéon-Sorbonne University（巴黎第一大学） ； University of Bucharest（布加勒斯特大学）
Comments Accepted at EMNLP Findings 2025
Time to Revist Exact Match
机构 * Language Technology Lab University of Cambridge（语言技术实验室剑桥大学） ； School of Computing Science University of Glasgow（计算科学学院格拉斯哥大学）
Comments Accepted for Findings of EMNLP 2025
Idiosyncratic Versus Normative Modeling of Atypical Speech Recognition: Dysarthric Case Studies
机构 * Stony Brook University（史蒂文尼森布鲁克大学） ； Vanderbilt University（范德比大学）
Comments Will appear in EMNLP 2025 Main Proceedings
OPEN-THEATRE: An Open-Source Toolkit for LLM-based Interactive Drama
机构 * UM–SJTU Joint Institute, Shanghai Jiao Tong University（上海交通大学与UM联合研究所） ； AGI Institute, School of Computer Science, Shanghai Jiao Tong University（上海交通大学人工智能研究所） ； Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering, Shanghai Jiao Tong University（上海教育委员会智能交互与认知工程重点实验室） ； Shanghai Key Laboratory of Trusted Data Circulation and Governance in Web3（上海Web3可信数据流通与治理重点实验室）
Comments Accepted by EMNLP 2025 demo
Decoding Uncertainty: The Impact of Decoding Strategies for Uncertainty Estimation in Large Language Models
机构 * Nara Institute of Science and Technology（奈良科学技术大学）
Comments Accepted at EMNLP 2025 Findings
ProtoVQA: An Adaptable Prototypical Framework for Explainable Fine-Grained Visual Question Answering
机构 * Dartmouth College（达特茅斯学院） ； Shandong University（山东大学） ； Harvard University（哈佛大学）
Comments Accepted to EMNLP 2025 Main Conference
When Big Models Train Small Ones: Label-Free Model Parity Alignment for Efficient Visual Question Answering using Small VLMs
机构 * Indian Institute of Technology Jodhpur（印度理工学院朱诺尔）
Comments Accepted to EMNLP (Main) 2025
LLMsPark: A Benchmark for Evaluating Large Language Models in Strategic Gaming Contexts
机构 * Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院） ； Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所） ； School of Software and Microelectronics, Peking University（北京大学软件与微电子学院） ； School of Computer Science and Engineering, Northeastern University（东北大学计算机科学与工程学院） ； Tongji University（同济大学） ； AIR, Tsinghua University（清华大学人工智能研究院） ； BAAI（百度人工智能研究院）
Comments Accepted by EMNLP 2025 Findings
SalaMAnder: Shapley-based Mathematical Expression Attribution and Metric for Chain-of-Thought Reasoning
机构 * Shanghai Jiao Tong University（上海交通大学） ； Alibaba Cloud Computing（阿里云计算）
Comments accpeted by EMNLP 2025
Captioning for Text-Video Retrieval via Dual-Group Direct Preference Optimization
机构 * Korea University（韩国大学） ； Hanwha Vision（翰威英航） ； KAIST（韩国科学技术院）
Comments EMNLP 2025 Findings
EMO-RL: Emotion-Rule-Based Reinforcement Learning Enhanced Audio-Language Model for Generalized Speech Emotion Recognition
Comments Accepted by the Findings of 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP Findings 2025)
Walk and Read Less: Improving the Efficiency of Vision-and-Language Navigation via Tuning-Free Multimodal Token Pruning
机构 * Boston University（波士顿大学）
Comments Accepted to EMNLP 2025. Data and code to be released at https://github.com/wdqin/VLN-NAP
Pun Unintended: LLMs and the Illusion of Humor Understanding
机构 * Dept of Environmental Sciences, Informatics and Statistics, Ca’ Foscari University of Venice（环境科学、信息与统计系，威尼斯卡弗西大学） ； School of Computer Science and Informatics, Cardiff University（计算机科学与信息学院，卡迪夫大学）
Comments Accepted to EMNLP 2025 Main Conference
The Good, the Bad and the Constructive: Automatically Measuring Peer Review's Utility for Authors
机构 * NLP Department, Mohamed Bin Zayed University of Artificial Intelligence（马尔代夫比兹艾兹大学人工智能学院自然语言处理系） ； Ubiquitous Knowledge Processing Lab, Department of Computer Science（计算机科学系通用知识处理实验室） ； Hessian Center for AI (hessian.AI), TU Darmstadt（图尔努尔德马斯特大学海斯塞人工智能中心）
Comments EMNLP 2025 Main
InMind: Evaluating LLMs in Capturing and Applying Individual Human Reasoning Styles
机构 * Nankai University（南开大学） ； Shanghai Innovation Institute（上海创新研究院） ； Shanghai AI Laboratory（上海人工智能实验室） ； Fudan University（复旦大学） ； Johns Hopkins University（约翰霍普金斯大学） ； Wuhan University（武汉大学） ； University of Science and Technology of China（中国科学技术大学）
Comments EMNLP 2025 MainConference
A Survey of Cognitive Distortion Detection and Classification in NLP
机构 * Department of Informatics, King’s College London（信息学院，伦敦国王学院）
Comments Camera-ready version to appear in EMNLP Findings 2025
The Missing Parts: Augmenting Fact Verification with Half-Truth Detection
机构 * School of Computing, National University of Singapore（计算学院，新加坡国立大学）
Comments Accepted by EMNLP 2025
Spec-VLA: Speculative Decoding for Vision-Language-Action Models with Relaxed Acceptance
机构 * NICS-EFC Lab, Department of Electronic Engineering, Tsinghua University（清华大学电子工程系NICS-EFC实验室） ； Department of Computer and Information Science, University of Macau（澳门大学计算机与信息科学系） ； Infinigence AI ； Tsinghua University（清华大学） ； Zhongguancun Academy（中关村学院）
Comments 13 pages, 5 figures, Accepted by EMNLP 2025 (main conference)
Dissecting Persona-Driven Reasoning in Language Models via Activation Patching
机构 * Independent（独立研究者）
Comments EMNLP (Findings) 2025
Latent Inter-User Difference Modeling for LLM Personalization
机构 * National University of Singapore（新加坡国立大学） ； University of Science and Technology of China（中国科学技术大学） ； The Chinese University of Hong Kong（香港中文大学）
Comments 2025 EMNLP Main Conference (Oral)
A Similarity Measure for Comparing Conversational Dynamics
机构 * Cornell University（康奈尔大学）
Comments Proceedings of EMNLP 2025 (Findings). Code and demos available in ConvoKit (https://convokit.cornell.edu/)
Journalism-Guided Agentic In-Context Learning for News Stance Detection
Comments EMNLP 2025 (24 pages)
PDFMathTranslate: Scientific Document Translation Preserving Layouts
机构 * National University of Singapore（新加坡国立大学） ； Tsinghua University（清华大学） ； University of Chinese Academy of Sciences（中国科学院大学） ； Funstory.ai Limited（Funstory.ai有限公司）
Comments 7 pages, 4 figures, EMNLP 2025 System Demonstration
GuiLoMo: Allocating Expert Number and Rank for LoRA-MoE via Bilevel Optimization with GuidedSelection Vectors
机构 * The University of Hong Kong（香港大学） ； Peking University（北京大学） ； Beijing University of Posts and Telecommunications（北京邮电大学） ； University of California, Los Angeles（加州大学洛杉矶分校） ； National Science Library, Chinese Academy of Sciences（中国科学院国家科学图书馆）
Comments Accepted by EMNLP 2025
Learning to Align: Addressing Character Frequency Distribution Shifts in Handwritten Text Recognition
机构 * Stony Brook University（石英溪大学） ； Archimedes, Athena Research Center（阿基米德研究所） ； Athens University of Economics and Business（雅典经济与商业大学）
Comments EMNLP 2025 Findings, 18 pages, 10 figures, 11 tables
From Chat Logs to Collective Insights: Aggregative Question Answering
机构 * University of Waterloo（滑铁卢大学） ； Cornell University（康奈尔大学）
Journal ref EMNLP 2025
Does quantization affect models' performance on long-context tasks?
Comments to appear in EMNLP 2025
BAGELS: Benchmarking the Automated Generation and Extraction of Limitations from Scholarly Text
机构 * Northern Illinois University（北伊利诺伊大学） ； University of North Texas（北德克萨斯大学）
Comments 16 pages, 8 figures, Accepted to the 2025 Conference on Empirical Methods in Natural Language Processing (Findings)
