语言大模型 / LLM

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） 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

Generalized Decoupled Learning for Enhancing Open-Vocabulary Dense Perception

Junjie Wang, Keyu Chen, Yulin Li, Bin Chen, Hengshuang Zhao, Xiaojuan Qi, Zhuotao Tian

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

Comments arXiv admin note: text overlap with arXiv:2505.04410

Approaching the Source of Symbol Grounding with Confluent Reductions of Abstract Meaning Representation Directed Graphs

Nicolas Goulet, Alexandre Blondin Massé, Moussa Abdendi

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

Hallucination vs interpretation: rethinking accuracy and precision in AI-assisted data extraction for knowledge synthesis

Xi Long, Christy Boscardin, Lauren A. Maggio, Joseph A. Costello, Ralph Gonzales, Rasmyah Hammoudeh, Ki Lai, Yoon Soo Park, Brian C. Gin

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

Provably Transformers Harness Multi-Concept Word Semantics for Efficient In-Context Learning

Dake Bu, Wei Huang, Andi Han, Atsushi Nitanda, Taiji Suzuki, Qingfu Zhang, Hau-San Wong

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

Comments Accepted by the 38th Conference on Neural Information Processing Systems (NeurIPS 2024)

Activation Steering for Bias Mitigation: An Interpretable Approach to Safer LLMs

Shivam Dubey

机构 * Indian Institute of Technology Madras（印度理工学院马德拉斯学院） 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.AI

MultiRAG: A Knowledge-guided Framework for Mitigating Hallucination in Multi-source Retrieval Augmented Generation

Wenlong Wu, Haofen Wang, Bohan Li, Peixuan Huang, Xinzhe Zhao, Lei Liang

机构 * 1 College of Artificial Intelligence, Nanjing University of Aeronautics ； Astronautics, Key Laboratory of Brain-Machine Intelligence Technology, Ministry of Education ； 2 College of Design \& Innovation, Tongji University ； 3 Key Laboratory of Intelligent Decision ； 4 Collaborative Innovation Center of Novel Software Technology 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

Comments Accepted by ICDE 2025 Research Paper

Journal ref In 2025 IEEE 41st International Conference on Data Engineering (ICDE), Hong Kong, 2025, pp. 3070-3083

Hide or Highlight: Understanding the Impact of Factuality Expression on User Trust

Hyo Jin Do, Werner Geyer

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

Comments 17 pages, 3 figures, To be published in Proceedings of the 8th AAAI/ACM Conference on AI, Ethics, and Society (AIES 2025)

Improving Factuality for Dialogue Response Generation via Graph-Based Knowledge Augmentation

Xiangyan Chen, Yujian Gan, Yimeng Gu, Matthew Purver

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

The Xeno Sutra: Can Meaning and Value be Ascribed to an AI-Generated "Sacred" Text?

Murray Shanahan, Tara Das, Robert Thurman

机构 * Imperial College London（帝国理工学院伦敦分校） ； University of London（伦敦大学） ； School of Advanced Study, University of London（伦敦大学高级研究学院） ； Columbia University（哥伦比亚大学） 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.AI

Modality Bias in LVLMs: Analyzing and Mitigating Object Hallucination via Attention Lens

Haohan Zheng, Zhenguo Zhang

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

Generative AI as a Pillar for Predicting 2D and 3D Wildfire Spread: Beyond Physics-Based Models and Traditional Deep Learning

Haowen Xu, Sisi Zlatanova, Ruiyu Liang, Ismet Canbulat

机构 * GRID, School of Built Environment, UNSW Sydney, NSW 2052 Australia（GRID，环境建筑学院，新南威尔士大学悉尼分校） ； School of Minerals and Energy Resources Engineering, UNSW Sydney, NSW 2052 Australia（矿物与能源资源工程学院，新南威尔士大学悉尼分校） 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.AI

CLIPTime: Time-Aware Multimodal Representation Learning from Images and Text

Anju Rani, Daniel Ortiz-Arroyo, Petar Durdevic

机构 * Department of Energy Technology（能源技术系） ； Aalborg University（奥尔堡大学） 纠错

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

Comments 11 pages, 8 figures

Can sparse autoencoders make sense of gene expression latent variable models?

Viktoria Schuster

机构 * Eric and Wendy Schmidt Center, Broad Institute of MIT and Harvard（埃里克和文迪斯中心，哈佛-麻省理工Broad研究所） ； Department of Computer Science, University of Copenhagen（哥本哈根大学计算机科学系） 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.LG

Comments 8 pages, 3 figures

On the Role of Discrete Representation in Sparse Mixture of Experts

Giang Do, Kha Pham, Hung Le, Truyen Tran

机构 * Applied Artificial Intelligence Institute (A2I2)（应用人工智能研究所） 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.LG

Comments 17 pages

Ta-G-T: Subjectivity Capture in Table to Text Generation via RDF Graphs

Ronak Upasham, Tathagata Dey, Pushpak Bhattacharyya

机构 * Department of Computer Science and Engineering（计算机科学与工程系） ； Indian Institute of Technology Bombay（印度理工学院孟买） 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

VitaGlyph: Vitalizing Artistic Typography with Flexible Dual-branch Diffusion Models

Kailai Feng, Yabo Zhang, Haodong Yu, Zhilong Ji, Jinfeng Bai, Hongzhi Zhang, Wangmeng Zuo

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

Comments https://github.com/Carlofkl/VitaGlyph

Multilingual Performance of a Multimodal Artificial Intelligence System on Multisubject Physics Concept Inventories

Gerd Kortemeyer, Marina Babayeva, Giulia Polverini, Ralf Widenhorn, Bor Gregorcic

机构 * AI Center, ETH Zurich（ETH Zurich人工智能中心） ； Michigan State University（密歇根州立大学） ； Charles University（查理大学） ； Uppsala University（乌普萨拉大学） ； Portland State University（波特兰州立大学） 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.AI

Journal ref Phys. Rev. Phys. Educ. Res. 21, 020101 (2025)

DLaVA: Document Language and Vision Assistant for Answer Localization with Enhanced Interpretability and Trustworthiness

Ahmad Mohammadshirazi, Pinaki Prasad Guha Neogi, Ser-Nam Lim, Rajiv Ramnath

机构 * Department of Computer Science（计算机科学系） ； Engineering, Ohio State University, Ohio, US（工程系，俄亥俄州立大学，俄亥俄，美国） ； Department of Computer Science, University of Central Florida, Florida, US（计算机科学系，中央佛罗里达大学，佛罗里达，美国） 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.AI

Agent Exchange: Shaping the Future of AI Agent Economics

Yingxuan Yang, Ying Wen, Jun Wang, Weinan Zhang

机构 * Shanghai Jiao Tong University（上海交通大学） ； University College London（伦敦大学学院） ； Shanghai Innovation Institute（上海创新研究院） 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.AI

Transformers Don't Need LayerNorm at Inference Time: Scaling LayerNorm Removal to GPT-2 XL and the Implications for Mechanistic Interpretability

Luca Baroni, Galvin Khara, Joachim Schaeffer, Marat Subkhankulov, Stefan Heimersheim

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

Uncertainty-aware Reward Design Process

Yang Yang, Xiaolu Zhou, Bosong Ding, Miao Xin

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

Comments 34 pages, 9 figures

Positional Bias in Binary Question Answering: How Uncertainty Shapes Model Preferences

Tiziano Labruna, Simone Gallo, Giovanni Da San Martino

机构 * University of Padova（帕多瓦大学） ； CNR-ISTI（意大利国家研究委员会-ISTI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

Performance of LLMs on Stochastic Modeling Operations Research Problems: From Theory to Practice

Akshit Kumar, Tianyi Peng, Yuhang Wu, Assaf Zeevi

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

Do LLMs Dream of Discrete Algorithms?

Claudionor Coelho, Yanen Li, Philip Tee

机构 * Zscaler Inc（Zscaler公司） ； ECE Department, Santa Clara University（圣克拉拉大学电子工程与计算机科学系） ； Department of Informatics, University of Sussex（苏塞克斯大学信息学院） ； The Beyond Center for Fundamental Science, Arizona State University（亚利桑那州立大学基础科学中心） 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.LG

Query as Test: An Intelligent Driving Test and Data Storage Method for Integrated Cockpit-Vehicle-Road Scenarios

Shengyue Yao, Runqing Guo, Yangyang Qin, Miangbing Meng, Jipeng Cao, Yilun Lin, Yisheng Lv, Fei-Yue Wang

机构 * Peking University International Innovation Center, Lin-gang Special Area (PKU-IICSH)（北京大学国际创新中心，临港特殊区域（PKU-IICSH）） ； Onesyn (Shanghai) Technology Co., Ltd（上海奥森科技有限公司） ； CATARC Automotive Technology(Shanghai) Co.,Ltd（CATARC汽车技术（上海）有限公司） ； ZEEKR Intelligent Technology Holding Limited（ZEKR智能技术控股有限公司） ； Department of Automation, Tsinghua University（清华大学自动化系） ； State Key Laboratory for Management and Control of Complex Systems, Chinese Academy of Sciences（复杂系统管理与控制国家重点实验室，中国科学院） ； Macao Institute of Systems Engineering, Macau University of Science and Technology（澳门系统工程研究院，澳门科技大学） 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.AI

Comments Submitted to IEEE Transaction on Vehicular Technology

Bilinear MLPs enable weight-based mechanistic interpretability

Michael T. Pearce, Thomas Dooms, Alice Rigg, Jose M. Oramas, Lee Sharkey

机构 * University of Antwerp（安特卫普大学） ； sqIRL/IDLab ； Apollo Research 纠错

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

Comments Accepted to ICLR'25

Abstract Meaning Representation for Hospital Discharge Summarization

Paul Landes, Sitara Rao, Aaron Jeremy Chaise, Barbara Di Eugenio

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

Chatting with Papers: A Hybrid Approach Using LLMs and Knowledge Graphs

Vyacheslav Tykhonov, Han Yang, Philipp Mayr, Jetze Touber, Andrea Scharnhorst

机构 * Networked Services, Royal Netherlands Academy of Arts（艺术与科学皇家荷兰科学院网络服务） ； Sciences (DANS-KNAW), 2593 HW The Hague, Anna van Saksenlaan 51, The Netherlands（科学（DANS-KNAW），荷兰海牙安娜·范·萨克森兰恩51号） ； GESIS -- Leibniz-Institute for the Social Sciences, Cologne, Germany（莱布尼茨社会科学研究所，科隆，德国） 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.AI

Comments 10 pages, 3 figures, Accepted at Joint Workshop of the 5th AI + Informetrics (AII) and the 6th Extraction and Evaluation of Knowledge Entities from Scientific Documents (EEKE)

Forecast-Then-Optimize Deep Learning Methods

Jinhang Jiang, Nan Wu, Ben Liu, Mei Feng, Xin Ji, Karthik Srinivasan

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

Comments 44 pages, 2 figures