语言大模型 / LLM

From Concepts to Components: Concept-Agnostic Attention Module Discovery in Transformers

Jingtong Su, Julia Kempe, Karen Ullrich

机构 * NYU（纽约大学） ； Meta AI ； FAIR 纠错

专题命中 知识编辑与模型理解 ：LLM(abstract)；post-training(abstract)；分类 cs.CL、cs.AI、cs.LG

MalGuard: Towards Real-Time, Accurate, and Actionable Detection of Malicious Packages in PyPI Ecosystem

Xingan Gao, Xiaobing Sun, Sicong Cao, Kaifeng Huang, Di Wu, Xiaolei Liu, Xingwei Lin, Yang Xiang

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

SAE-V: Interpreting Multimodal Models for Enhanced Alignment

Hantao Lou, Changye Li, Jiaming Ji, Yaodong Yang

机构 * Institute for AI, Peking University, Beijing, China（人工智能研究院，北京大学，北京，中国） ； State Key Laboratory of General Artificial Intelligence, Institute for AI, Peking University, Beijing, China（通用人工智能国家重点实验室，人工智能研究院，北京大学，北京，中国） 纠错

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

Comments 17 pages, 13 figures

Why Uncertainty Estimation Methods Fall Short in RAG: An Axiomatic Analysis

Heydar Soudani, Evangelos Kanoulas, Faegheh Hasibi

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

Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models

Femi Bello, Anubrata Das, Fanzhi Zeng, Fangcong Yin, Liu Leqi

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

COMPKE: Complex Question Answering under Knowledge Editing

Keyuan Cheng, Zijian Kan, Zhixian He, Zhuoran Zhang, Muhammad Asif Ali, Ke Xu, Lijie Hu, Di Wang

机构 * Provable Responsible AI and Data Analytics (PRADA) Lab（可证明负责任的人工智能与数据分析实验室） ； King Abdullah University of Science and Technology（卡布斯大学） ； Peking University（北京大学） ； South China University of Technology（华南理工大学） ； Sun Yat-sen University（中山大学） 纠错

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

Comments Accepted by ACL 2025 Findings

HD-NDEs: Neural Differential Equations for Hallucination Detection in LLMs

Qing Li, Jiahui Geng, Zongxiong Chen, Derui Zhu, Yuxia Wang, Congbo Ma, Chenyang Lyu, Fakhri Karray

机构 * Mohamed bin Zayed University of Artificial Intelligence（莫扎德·本·泽亚德人工智能大学） ； Fraunhofer Institute for Open Communication Systems（弗劳恩霍夫开放通信系统研究所） ； Technical University of Munich（慕尼黑技术大学） ； New York University Abu Dhabi（纽约大学阿布扎克分校） ； Alibaba International Digital Commerce（阿里巴巴国际数字商业） 纠错

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

An AI-powered Knowledge Hub for Potato Functional Genomics

Jia Yuxin, Li Jinye, Jia Yudong, Li Futing, Su Xiaoqi, Luo Jilin, Dong Yarui, Sun Chunyan, Cui Qinghan, Wang Li, Li Axiu, Shang Yi, Zhu Yujuan, Huang Sanwen

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

Comments 11 pages, 4 figures

Anchored Answers: Unravelling Positional Bias in GPT-2's Multiple-Choice Questions

Ruizhe Li, Yanjun Gao

机构 * Department of Computing Science, University of Aberdeen, UK（阿伯丁大学计算机科学系） ； University of Colorado Anschutz Medical Campus, United States（科罗拉多大学安舒茨医学校区） 纠错

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

Comments ACL 2025 Findings

Information Structure in Mappings: An Approach to Learning, Representation, and Generalisation

Henry Conklin

机构 * Henry Coxe Conklin（独立研究者） 纠错

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

Comments PhD Thesis, 204 pages; entropy estimation discussed from p.94

Can LLMs Deceive CLIP? Benchmarking Adversarial Compositionality of Pre-trained Multimodal Representation via Text Updates

Jaewoo Ahn, Heeseung Yun, Dayoon Ko, Gunhee Kim

机构 * Seoul National University（首尔国立大学） 纠错

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

Comments ACL 2025 Main. Code is released at https://vision.snu.ac.kr/projects/mac

Mechanistic Interpretability of GPT-like Models on Summarization Tasks

Anurag Mishra

机构 * Rochester Institute of Technology（罗切斯特技术研究所） 纠错

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

Comments 8 pages (6 content + 2 references/appendix), 6 figures, 2 tables; under review for the ACL 2025 Student Research Workshop

Lifelong Knowledge Editing requires Better Regularization

Akshat Gupta, Phudish Prateepamornkul, Maochuan Lu, Ahmed Alaa, Thomas Hartvigsen, Gopala Anumanchipalli

机构 * UC Berkeley（加州大学伯克利分校） ； SCB DataX（SCB数据科技） ； University of Virginia（弗吉尼亚大学） 纠错

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

CRCE: Coreference-Retention Concept Erasure in Text-to-Image Diffusion Models

Yuyang Xue, Edward Moroshko, Feng Chen, Jingyu Sun, Steven McDonagh, Sotirios A. Tsaftaris

机构 * School of Engineering University of Edinburgh（工程学院 苏格兰爱丁堡大学） 纠错

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

Mitigating Hallucination in VideoLLMs via Temporal-Aware Activation Engineering

Jianfeng Cai, Wengang Zhou, Zongmeng Zhang, Jiale Hong, Nianji Zhan, Houqiang Li

机构 * University of Science and Technology of China（中国科学技术大学） ； Shanghai Jiaotong University（上海交通大学） ； Merchants Union Consumer Finance Company Limited 纠错

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

A Note on Semantic Diffusion

Alexander P. Ryjov, Alina A. Egorova

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

Comments 8 figures

Beyond Single Concept Vector: Modeling Concept Subspace in LLMs with Gaussian Distribution

Haiyan Zhao, Heng Zhao, Bo Shen, Ali Payani, Fan Yang, Mengnan Du

机构 * New Jersey Institute of Technology（新泽西理工学院） ； Wake Forest University（威克森林大学） ； Cisco Research（思科研究） 纠错

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

Comments Accepted by ICLR 2025

JailbreakLens: Interpreting Jailbreak Mechanism in the Lens of Representation and Circuit

Zeqing He, Zhibo Wang, Zhixuan Chu, Huiyu Xu, Wenhui Zhang, Qinglong Wang, Rui Zheng

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

Comments 17 pages, 11 figures

Scaling sparse feature circuit finding for in-context learning

Dmitrii Kharlapenko, Stepan Shabalin, Fazl Barez, Arthur Conmy, Neel Nanda

机构 * University of Oxford, UK（牛津大学） ； ETH Zurich, Switzerland（苏黎世联邦理工学院） ； Georgia Institute of Technology, US（佐治亚理工学院） 纠错

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

Randomness, Not Representation: The Unreliability of Evaluating Cultural Alignment in LLMs

Ariba Khan, Stephen Casper, Dylan Hadfield-Menell

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

Towards Interpretable Soft Prompts

Oam Patel, Jason Wang, Nikhil Shivakumar Nayak, Suraj Srinivas, Himabindu Lakkaraju

专题命中 知识编辑与模型理解 ：LLM(abstract)；prompting(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 9 pages, 8 figures

SUNAR: Semantic Uncertainty based Neighborhood Aware Retrieval for Complex QA

V Venktesh, Mandeep Rathee, Avishek Anand

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

Comments Accepted at NAACL 2025 Main Conference

MASH-VLM: Mitigating Action-Scene Hallucination in Video-LLMs through Disentangled Spatial-Temporal Representations

Kyungho Bae, Jinhyung Kim, Sihaeng Lee, Soonyoung Lee, Gunhee Lee, Jinwoo Choi

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

Comments Accepted for CVPR 2025

GASP: Unifying Geometric and Semantic Self-Supervised Pre-training for Autonomous Driving

William Ljungbergh, Adam Lilja, Adam Tonderski. Arvid Laveno Ling, Carl Lindström, Willem Verbeke, Junsheng Fu, Christoffer Petersson, Lars Hammarstrand, Michael Felsberg

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

Transformers Struggle to Learn to Search

Abulhair Saparov, Srushti Pawar, Shreyas Pimpalgaonkar, Nitish Joshi, Richard Yuanzhe Pang, Vishakh Padmakumar, Seyed Mehran Kazemi, Najoung Kim, He He

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

Comments Published as a conference paper at ICLR 2025

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

Zhengxuan Wu, Aryaman Arora, Atticus Geiger, Zheng Wang, Jing Huang, Dan Jurafsky, Christopher D. Manning, Christopher Potts

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

Representation Engineering: A Top-Down Approach to AI Transparency

Andy Zou, Long Phan, Sarah Chen, James Campbell, Phillip Guo, Richard Ren, Alexander Pan, Xuwang Yin, Mantas Mazeika, Ann-Kathrin Dombrowski, Shashwat Goel, Nathaniel Li, Michael J. Byun, Zifan Wang, Alex Mallen, Steven Basart, Sanmi Koyejo, Dawn Song, Matt Fredrikson, J. Zico Kolter, Dan Hendrycks

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

Comments Code is available at https://github.com/andyzoujm/representation-engineering

Towards General Visual-Linguistic Face Forgery Detection(V2)

Ke Sun, Shen Chen, Taiping Yao, Ziyin Zhou, Jiayi Ji, Xiaoshuai Sun, Chia-Wen Lin, Rongrong Ji

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

Comments 8 pages, 5 figures, Accpet by CVPR2025

MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation

Chenxi Wang, Xiang Chen, Ningyu Zhang, Bozhong Tian, Haoming Xu, Shumin Deng, Huajun Chen

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

Comments ICLR 2025

An explainable transformer circuit for compositional generalization

Cheng Tang, Brenden Lake, Mehrdad Jazayeri

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