Rethinking the Role of Text Complexity in Language Model Pretraining
机构 * Samsung R&D Institute Philippines（三星菲律宾研发中心）
Comments Camera-ready version for BabyLM Workshop at EMNLP 2025
期刊&会议
EMNLP
Conference on Empirical Methods in Natural Language Processing · 会议 · Natural Language Processing
CEMTM: Contextual Embedding-based Multimodal Topic Modeling
Comments EMNLP 2025
Meta-Pretraining for Zero-Shot Cross-Lingual Named Entity Recognition in Low-Resource Philippine Languages
机构 * University of Cambridge（剑桥大学）
Comments Accepted (poster) to 5th Workshop on Multilingual Representation Learning at EMNLP 2025
LPI-RIT at LeWiDi-2025: Improving Distributional Predictions via Metadata and Loss Reweighting with DisCo
机构 * Rochester Institute of Technology（罗切斯特理工学院）
Comments To appear in Proceedings of the EMNLP 2025 Workshop on Learning with Disagreements (LeWiDi)
XAutoLM: Efficient Fine-Tuning of Language Models via Meta-Learning and AutoML
机构 * University of Alicante（阿利坎特大学） ； University of Havana（哈瓦那大学） ； University of Lancaster（兰卡斯特大学）
Comments 18 pages, 10 figures, 7 tables. Preprint. Accepted at EMNLP 2025
User Feedback in Human-LLM Dialogues: A Lens to Understand Users But Noisy as a Learning Signal
机构 * New York University（纽约大学）
Comments EMNLP camera-ready
C3: A Bilingual Benchmark for Spoken Dialogue Models Exploring Challenges in Complex Conversations
机构 * Peking University（北京大学） ； LIGHTSPEED
Comments EMNLP 2025 main; Project Page: https://step-out.github.io/C3-web/
TopicAttack: An Indirect Prompt Injection Attack via Topic Transition
Comments EMNLP 2025
The Security Threat of Compressed Projectors in Large Vision-Language Models
机构 * Department of Electronic Engineering, Tsinghua University（清华大学电子工程系） ； Large Language Model Department, Tencent（腾讯大语言模型部门） ； School of Computer and Communication Engineering, University of Science and Technology Beijing（北京科技大学计算机与通信工程学院） ； Faculty of Science and Technology, University of Macau（澳门大学科技学院）
Comments Accepted by EMNLP 2025 findings
LegalSearchLM: Rethinking Legal Case Retrieval as Legal Elements Generation
Comments EMNLP 2025 Main Conference
Scientific Paper Retrieval with LLM-Guided Semantic-Based Ranking
机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； Korea University（韩国大学）
Comments Accepted to EMNLP 2025 Findings
From Word to World: Evaluate and Mitigate Culture Bias in LLMs via Word Association Test
机构 * The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳）） ； Shenzhen Research Institute of Big Data（深圳大数据研究院）
Comments Cultural Analysis, Cultural Alignment, Word Association Test, Large Language Models. Accepted by EMNLP 2025 (Oral)
Can Large Language Models Outperform Non-Experts in Poetry Evaluation? A Comparative Study Using the Consensual Assessment Technique
Comments 18 pages, 3 figures. Accepted for publication at the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Detecting LLM Hallucination Through Layer-wise Information Deficiency: Analysis of Ambiguous Prompts and Unanswerable Questions
机构 * University of Oxford（牛津大学）
Comments Accepted to EMNLP(main)2025
Searching for the Most Human-like Emergent Language
机构 * Language Technologies Institute Carnegie Mellon University（语言技术研究所卡内基梅隆大学）
Comments Accepted for publication at the 2025 Conference on Empirical Methods in Natural Language Processing; 19 pages, 12 figures
Morpheme Induction for Emergent Language
机构 * Language Technologies Institute Carnegie Mellon University（语言技术研究所卡内基梅隆大学）
Comments Accepted for publication at the 2025 Conference on Empirical Methods in Natural Language Processing; 16 pages, 4 figures
XTRA: Cross-Lingual Topic Modeling with Topic and Representation Alignments
机构 * Hanoi University of Science and Technology（河内科学技术大学） ； University of Monash（墨尔本大学）
Comments 2025 EMNLP Findings
Self-Improvement in Multimodal Large Language Models: A Survey
机构 * The University of Texas at Dallas（德克萨斯大学达拉斯分校） ； University of Toronto（多伦多大学） ； University of Notre Dame（诺特丹大学） ； Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）
Comments EMNLP 2025
SoT: Structured-of-Thought Prompting Guides Multilingual Reasoning in Large Language Models
机构 * Key Laboratory of Big Data & Artificial Intelligence in Transportation (Beijing Jiaotong University), Ministry of Education（大数据与人工智能交通 key laboratory（北京交通大学）） ； School of Computer Science and Technology, Beijing Jiaotong University（计算机科学与技术学院（北京交通大学）） ； University of Montreal（蒙特利尔大学）
Comments EMNLP 2025 (findings)
AMANDA: Agentic Medical Knowledge Augmentation for Data-Efficient Medical Visual Question Answering
机构 * Northwestern University（西北大学） ； Microsoft（微软公司）
Comments EMNLP Findings
The Prompt Makes the Person(a): A Systematic Evaluation of Sociodemographic Persona Prompting for Large Language Models
机构 * University of Mannheim（曼海姆大学） ； GESIS - Leibniz Institute for the Social Sciences（莱布尼茨社会科学研究所） ； Complexity Science Hub Vienna（维也纳复杂科学中心）
Comments Accepted to EMNLP Findings 2025
A Survey of Pun Generation: Datasets, Evaluations and Methodologies
机构 * School of Computer Science, University of Auckland（奥克兰大学计算机科学学院） ； Singapore University of Technology and Design（新加坡科技设计大学） ； School of Electronic Engineering and Computer Science, Queen Mary University of London（伦敦女王学院电子工程与计算机科学学院）
Comments Accepted to EMNLP 2025 Findings
From Long Videos to Engaging Clips: A Human-Inspired Video Editing Framework with Multimodal Narrative Understanding
机构 * University of Science and Technology of China（中国科学技术大学） ； ByteDance China（字节跳动中国）
Comments Accepted by EMNLP 2025 Industry Track
Same evaluation, more tokens: On the effect of input length for machine translation evaluation using Large Language Models
Comments Accepted at EMNLP 2025 (Main Conference)
DatawiseAgent: A Notebook-Centric LLM Agent Framework for Adaptive and Robust Data Science Automation
机构 * National Engineering Research Center for Software Engineering, Peking University（软件工程国家工程研究中心，北京大学） ； School of Software & Microelectronics, Peking University（软件与微电子学院，北京大学） ； School of Computer Science, Peking University（计算机科学学院，北京大学） ； Xi’an Jiaotong University（西安交通大学） ； Institute of Basic Theory of Chinese Medicine, China Academy of Chinese Medical Sciences（中医基础理论研究所，中国中医科学院）
Comments The camera-ready version for EMNLP 2025 Main Conference
Primus: A Pioneering Collection of Open-Source Datasets for Cybersecurity LLM Training
机构 * AI Lab, TrendMicro（TrendMicro人工智能实验室）
Comments Accepted to EMNLP 2025
RACCooN: A Versatile Instructional Video Editing Framework with Auto-Generated Narratives
机构 * UNC Chapel Hill（北卡罗来纳大学教堂山分校） ； Nanyang Technological University（南洋理工大学）
Comments EMNLP 2025 main; The first two authors contribute equally. Project Page: https://raccoon-mllm-gen.github.io/
Interactive Training: Feedback-Driven Neural Network Optimization
机构 * University of Waterloo（滑铁卢大学） ； University of Wisconsin-Madison（威斯康星大学麦迪逊分校）
Comments EMNLP 2025 Demo
From Behavioral Performance to Internal Competence: Interpreting Vision-Language Models with VLM-Lens
Comments EMNLP 2025 System Demonstration | Code: https://github.com/compling-wat/vlm-lens
FlexDoc: Parameterized Sampling for Diverse Multilingual Synthetic Documents for Training Document Understanding Models
机构 * Oracle Corporation（Oracle公司）
Comments Accepted at EMNLP 2025
