语言大模型 / LLM - arXivDaily 专题

2406.05328 2025-08-19 cs.CL cs.LG 90%

FacLens: Transferable Probe for Foreseeing Non-Factuality in Fact-Seeking Question Answering of Large Language Models

Yanling Wang, Haoyang Li, Hao Zou, Jing Zhang, Xinlei He, Qi Li, Ke Xu

机构 * Zhipu AI（智谱AI）； Zhongguancun Laboratory（中关村实验室）； Renmin University of China（中国人民大学）； Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））； Tsinghua University（清华大学）

专题命中知识编辑与模型理解：large language model(title,abstract)；language model(title,abstract)；LLM(abstract)；分类 cs.CL、cs.LG

2508.12482 2025-08-19 cs.CL 88%

The Structural Sources of Verb Meaning Revisited: Large Language Models Display Syntactic Bootstrapping

Xiaomeng Zhu, R. Thomas McCoy, Robert Frank

机构 * Department of Linguistics, Yale University（语言学系，耶鲁大学）； Wu Tsai Institute, Yale University（吴 Tsai 院，耶鲁大学）

专题命中知识编辑与模型理解：large language model(title,abstract)；language model(title,abstract)；分类 cs.CL

2508.13152 2025-08-19 cs.CL cs.AI 86%

RepreGuard: Detecting LLM-Generated Text by Revealing Hidden Representation Patterns

Xin Chen, Junchao Wu, Shu Yang, Runzhe Zhan, Zeyu Wu, Ziyang Luo, Di Wang, Min Yang, Lidia S. Chao, Derek F. Wong

机构 * NLP（自然语言处理）； CT Lab, Department of Computer and Information Science, University of Macau（计算机与信息科学系，澳门大学）； Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences（深圳先进技术研究院，中国科学院）； Provable Responsible AI and Data Analytics Lab, KAUST（可证明责任AI与数据分析实验室，卡尔斯兰大学）； Hong Kong Baptist University（香港 Baptist 大学）

专题命中知识编辑与模型理解：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL、cs.AI

Comments Accepted to TACL 2025. This version is a pre-MIT Press publication version

2506.16550 2025-08-19 cs.LG stat.ML 83%

A Free Probabilistic Framework for Analyzing the Transformer-based Language Models

Swagatam Das

机构 * Electronics and Communication Sciences Unit, Indian Statistical Institute, Kolkata, India.（印度统计研究所电子与通信科学单元，加尔各答，印度）

专题命中知识编辑与模型理解：language model(title,abstract)；large language model(abstract)；分类 cs.LG

2504.07402 2025-08-19 cs.LG cs.AI 81%

LauraTSE: Target Speaker Extraction using Auto-Regressive Decoder-Only Language Models

Beilong Tang, Bang Zeng, Ming Li

机构 * Duke Kunshan University（杜克昆山大学）

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.AI、cs.LG

Comments 8 pages, 5 figure, accepted by 2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)

2508.12586 2025-08-19 cs.CV 78%

Foundation Model for Skeleton-Based Human Action Understanding

Hongsong Wang, Wanjiang Weng, Junbo Wang, Fang Zhao, Guo-Sen Xie, Xin Geng, Liang Wang

机构 * School of Computer Science and Engineering, Southeast University（东南大学计算机科学与工程学院）； Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications (Southeast University), Ministry of Education, China（新一代人工智能技术及交叉应用关键实验室（东南大学），中华人民共和国教育部，中国）； School of Software, Northwestern Polytechnical University（西北工业大学软件学院）； State Key Laboratory for Novel Software Technology and School of Intelligence Science and Technology, Nanjing University（新型软件技术国家重点实验室和南京大学智能科学与技术学院）； School of Computer Science and Engineering, Nanjing University of Science and Technology（南京理工大学计算机科学与工程学院）

专题命中知识编辑与模型理解：foundation model(title,abstract)

Comments Accepted by TPAMI, Code is available at: https://github.com/wengwanjiang/FoundSkelModel

2404.10357 2025-08-19 cs.CV 78%

Optimization of Prompt Learning via Multi-Knowledge Representation for Vision-Language Models

Enming Zhang, Bingke Zhu, Yingying Chen, Qinghai Miao, Ming Tang, Jinqiao Wang

机构 * School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）； Foundation Model Research Center, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所基础模型研究中心）； Wuhan AI Research（武汉人工智能研究所）； Peng Cheng Laboratory（鹏城实验室）

专题命中知识编辑与模型理解：language model(title,abstract)

2411.16252 2025-08-19 cs.CL 77%

NormXLogit: The Head-on-Top Never Lies

Sina Abbasi, Mohammad Reza Modarres, Mohammad Taher Pilehvar

机构 * Tehran Institute for Advanced Studies, Khatam University, Iran（泰赫兰高级研究院，卡坦大学，伊朗）； Cardiff University（卡迪夫大学）

专题命中知识编辑与模型理解：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments Added comparisons on computational efficiency, included experiments on a new dataset with an additional evaluation metric for classification tasks, expanded explanations and discussions in the experiments, and presented a worked example for alignment metrics computation

2508.11933 2025-08-19 cs.CL 77%

CAMF: Collaborative Adversarial Multi-agent Framework for Machine Generated Text Detection

Yue Wang, Liesheng Wei, Yuxiang Wang

机构 * Stanford University（斯坦福大学）； College of Information Technology（信息科技学院）； School of Business（商学院）

专题命中知识编辑与模型理解：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

2508.12803 2025-08-19 cs.CL 70%

When Alignment Hurts: Decoupling Representational Spaces in Multilingual Models

Ahmed Elshabrawy, Hour Kaing, Haiyue Song, Alham Fikri Aji, Hideki Tanaka, Masao Utiyama, Raj Dabre

机构 * MBZUAI（马克斯·普朗克人工智能研究所）； NICT, Japan（日本信息通信技术研究所）； IIT Madras（印度理工学院Madras分校）

专题命中知识编辑与模型理解：large language model(abstract)；language model(abstract)；分类 cs.CL

2506.12609 2025-08-19 cs.CV 50%

Not All Tokens and Heads Are Equally Important: Dual-Level Attention Intervention for Hallucination Mitigation

Lexiang Tang, Xianwei Zhuang, Bang Yang, Zhiyuan Hu, Hongxiang Li, Lu Ma, Jinghan Ru, Yuexian Zou

专题命中知识编辑与模型理解：language model(abstract)

2507.15227 2025-08-19 cs.CV 50%

Mammo-SAE: Interpreting Breast Cancer Concept Learning with Sparse Autoencoders

Krishna Kanth Nakka

机构 * institutetext: Bavaria, Germany（巴伐利亚，德国）

专题命中知识编辑与模型理解：foundation model(abstract)

Comments Accepted at Deep Breast Imaging workshop, MICCAI 2025

1. 知识编辑与模型理解 12 篇