大模型对齐与安全 - arXivDaily 专题

2502.16688 2025-02-25 cs.LG 57%

Analyzing Factors Influencing Driver Willingness to Accept Advanced Driver Assistance Systems

Hannah Musau, Nana Kankam Gyimah, Judith Mwakalonge, Gurcan Comert, Saidi Siuhi

机构 * South Carolina State University（南卡罗来纳州立大学）； North Carolina A&T State University（北卡罗来纳农工州立大学）

专题命中其他安全：safety(abstract)；分类 cs.LG

2408.10593 2025-02-25 cs.CL cs.CV 57%

An Efficient Sign Language Translation Using Spatial Configuration and Motion Dynamics with LLMs

Eui Jun Hwang, Sukmin Cho, Junmyeong Lee, Jong C. Park

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments Accepted to NAACL 2025 main

2406.09288 2025-02-25 cs.LG 57%

Large Language Model as a Teacher for Zero-shot Tagging at Extreme Scales

Jinbin Zhang, Nasib Ullah, Rohit Babbar

机构 * Aalto University（阿尔托大学）； University of Bath（巴斯大学）

专题命中其他安全：alignment(abstract)；分类 cs.LG

2502.15924 2025-02-25 cs.CL 57%

Improving Consistency in Large Language Models through Chain of Guidance

Harsh Raj, Vipul Gupta, Domenic Rosati, Subhabrata Majumdar

机构 * Northeastern University（东北大学）； Pennsylvania State University（宾夕法尼亚州立大学）； Dalhousie University（达尔豪斯大学）； Vijil（维吉尔科技）

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments Accepted at Transactions of Machine Learning Research (TMLR) 2025

2411.12980 2025-02-25 cs.CV cs.AI 57%

LaVida Drive: Vision-Text Interaction VLM for Autonomous Driving with Token Selection, Recovery and Enhancement

Siwen Jiao, Yangyi Fang, Baoyun Peng, Wangqun Chen, Bharadwaj Veeravalli

机构 * National University of Singapore（新加坡国立大学）； Tsinghua University（清华大学）； Agency for Science, Technology and Research, Singapore（新加坡科技研究局）； Advanced Institute of Big Data, Beijing（北京大数据研究院）

专题命中其他安全：alignment(abstract)；分类 cs.AI

2502.15412 2025-02-24 cs.CL 57%

Textual-to-Visual Iterative Self-Verification for Slide Generation

Yunqing Xu, Xinbei Ma, Jiyang Qiu, Hai Zhao

机构 * Shanghai Jiao Tong University（上海交通大学）

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments Work in progress

2502.14420 2025-02-24 cs.RO cs.CV cs.LG 57%

ChatVLA: Unified Multimodal Understanding and Robot Control with Vision-Language-Action Model

Zhongyi Zhou, Yichen Zhu, Minjie Zhu, Junjie Wen, Ning Liu, Zhiyuan Xu, Weibin Meng, Ran Cheng, Yaxin Peng, Chaomin Shen, Feifei Feng

机构 * Midea Group（美的集团）； East China Normal University（华东师范大学）； Shanghai University（上海大学）； Beijing Innovation Center of Humanoid Robotics（北京人形机器人创新中心）； Tsinghua University（清华大学）

专题命中其他安全：alignment(abstract)；分类 cs.LG

2502.14735 2025-02-21 cs.IR cs.AI 57%

EAGER-LLM: Enhancing Large Language Models as Recommenders through Exogenous Behavior-Semantic Integration

Minjie Hong, Yan Xia, Zehan Wang, Jieming Zhu, Ye Wang, Sihang Cai, Xiaoda Yang, Quanyu Dai, Zhenhua Dong, Zhimeng Zhang, Zhou Zhao

机构 * Zhejiang University（浙江大学）； Huawei Noah’s Ark Lab（华为诺亚方舟实验室）

专题命中其他安全：alignment(abstract)；分类 cs.AI

Comments 9 pages, 6 figures, accpeted by WWW 2025

2502.14571 2025-02-21 cs.LG cs.CE 57%

Predicting Filter Medium Performances in Chamber Filter Presses with Digital Twins Using Neural Network Technologies

Dennis Teutscher, Tyll Weber-Carstanjen, Stephan Simonis, Mathias J. Krause

机构 * Karlsruhe Institute of Technology（卡尔斯鲁厄理工学院）

专题命中其他安全：alignment(abstract)；分类 cs.LG

2502.14182 2025-02-21 cs.CR cs.LG 57%

Multi-Faceted Studies on Data Poisoning can Advance LLM Development

Pengfei He, Yue Xing, Han Xu, Zhen Xiang, Jiliang Tang

机构 * Michigan State University（密歇根州立大学）； University of Arizona（亚利桑那大学）； University of Georgia（佐治亚大学）

专题命中其他安全：safety(abstract)；分类 cs.LG

2502.10916 2025-02-21 cs.CL cs.IR 57%

An Open-Source Web-Based Tool for Evaluating Open-Source Large Language Models Leveraging Information Retrieval from Custom Documents

Godfrey I

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments 19 pages, 1 figure, 6 tables

2412.02861 2025-02-21 stat.ML cs.LG 57%

An Information-Theoretic Analysis of Thompson Sampling for Logistic Bandits

Amaury Gouverneur, Borja Rodríguez-Gálvez, Tobias J. Oechtering, Mikael Skoglund

机构 * KTH Royal Institute of Technology（瑞典皇家理工学院）

专题命中其他安全：alignment(abstract)；分类 cs.LG

Comments 21 pages, under review

2308.08469 2025-02-21 cs.LG 57%

LLM4TS: Aligning Pre-Trained LLMs as Data-Efficient Time-Series Forecasters

Ching Chang, Wei-Yao Wang, Wen-Chih Peng, Tien-Fu Chen

机构 * Institute for Clarity in Documentation（文献清晰研究所）； The Thørväld Group（索沃尔德集团）； Inria（法国国家信息与自动化研究所）； Rajiv Gandhi University（拉吉夫·甘地大学）； Tsinghua University（清华大学）； Palmer Research Laboratories（帕尔默研究实验室）； The Kumquat Consortium（柑橘联盟）

专题命中其他安全：alignment(abstract)；分类 cs.LG

Comments Accepted for publication in ACM Transactions on Intelligent Systems and Technology (TIST) 2025. The final published version will be available at https://doi.org/10.1145/3719207

2502.13476 2025-02-20 cs.AI cs.NI 57%

Integration of Agentic AI with 6G Networks for Mission-Critical Applications: Use-case and Challenges

Sunder Ali Khowaja, Kapal Dev, Muhammad Salman Pathan, Engin Zeydan, Merouane Debbah

机构 * School of Computing, Dublin City University（都柏林城市大学计算机学院）； ADAPT Centre（ADAPT中心）； Munster Technological University（芒斯特理工大学）； Centre Tecnològic de Telecomunicacions de Catalunya（加泰罗尼亚电信技术中心）； Khalifa University（哈利法大学）

专题命中其他安全：safety(abstract)；分类 cs.AI

Comments FEMA [https://www.fema.gov/openfema-data-page/disaster-declarations-summaries-v2] National Oceanic and Atmospheric Administration [https://www.ncdc.noaa.gov/stormevents/details.jsp] packages Pytorch [https://pytorch.org/] RLib [https://docs.ray.io/en/latest/rllib/index.html] Neo4j [https://neo4j.com/] Apache Kafka [https://kafka.apache.org/]

2502.13396 2025-02-20 cs.CL 57%

Prompting a Weighting Mechanism into LLM-as-a-Judge in Two-Step: A Case Study

Wenwen Xie, Gray Gwizdz, Dongji Feng

机构 * Databricks ； Gustavus Adolphus College（古斯塔夫斯·阿道弗斯学院）

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments 5 pages, 5 tables, 1 figure

2502.12948 2025-02-19 cs.CV cs.AI 57%

Fake It Till You Make It: Using Synthetic Data and Domain Knowledge for Improved Text-Based Learning for LGE Detection

Athira J Jacob, Puneet Sharma, Daniel Rueckert

专题命中其他安全：alignment(abstract)；分类 cs.AI

Comments Poster at Workshop on Large Language Models and Generative AI for Health at AAAI 2025

2502.11495 2025-02-18 cs.CL 57%

Balanced Multi-Factor In-Context Learning for Multilingual Large Language Models

Masahiro Kaneko, Alham Fikri Aji, Timothy Baldwin

机构 * MBZUAI（穆罕默德·本·扎耶德人工智能大学）

专题命中其他安全：alignment(abstract)；分类 cs.CL

2407.16724 2025-02-18 cs.CL 57%

Structure-aware Domain Knowledge Injection for Large Language Models

Kai Liu, Ze Chen, Zhihang Fu, Wei Zhang, Rongxin Jiang, Fan Zhou, Yaowu Chen, Yue Wu, Jieping Ye

机构 * Zhejiang University（浙江大学）； Alibaba Cloud（阿里云）

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments Preprint. Code is available at https://github.com/alibaba/struxgpt

2404.14750 2025-02-18 cs.CV cs.AI 57%

Grounded Knowledge-Enhanced Medical Vision-Language Pre-training for Chest X-Ray

Qiao Deng, Zhongzhen Huang, Yunqi Wang, Zhichuan Wang, Zhao Wang, Xiaofan Zhang, Qi Dou, Yeung Yu Hui, Edward S. Hui

机构 * The Chinese University of Hong Kong（香港中文大学）； CU Lab for AI in Radiology (CLAIR), The Chinese University of Hong Kong（香港中文大学放射学人工智能实验室（CLAIR））； Shanghai Jiao Tong University（上海交通大学）； Shanghai AI Laboratory（上海人工智能实验室）； China Unicom Global Limited（中国联通国际有限公司）

专题命中其他安全：alignment(abstract)；分类 cs.AI

2502.10140 2025-02-17 cs.CL 57%

Small Models, Big Impact: Efficient Corpus and Graph-Based Adaptation of Small Multilingual Language Models for Low-Resource Languages

Daniil Gurgurov, Ivan Vykopal, Josef van Genabith, Simon Ostermann

机构 * University of Saarland（萨尔大学）； Brno University of Technology（布尔诺理工大学）； German Research Center for Artificial Intelligence (DFKI)（德国人工智能研究中心）； Kempelen Institute of Intelligent Technologies (KInIT)（肯佩伦智能技术研究所）

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments Pre-print

2502.09218 2025-02-14 cs.LO cs.AI 57%

Data2Concept2Text: An Explainable Multilingual Framework for Data Analysis Narration

Flavio Bertini, Alessandro Dal Palù, Federica Zaglio, Francesco Fabiano, Andrea Formisano

机构 * University of Parma（帕尔马大学）； New Mexico State University（新墨西哥州立大学）； University of Udine（乌迪内大学）

专题命中其他安全：safety(abstract)；分类 cs.AI

Comments In Proceedings ICLP 2024, arXiv:2502.08453

Journal ref EPTCS 416, 2025, pp. 139-152

2502.09209 2025-02-14 cs.AI 57%

On LLM-generated Logic Programs and their Inference Execution Methods

Paul Tarau

机构 * University of North Texas（北得克萨斯大学）

专题命中其他安全：alignment(abstract)；分类 cs.AI

Comments In Proceedings ICLP 2024, arXiv:2502.08453

Journal ref EPTCS 416, 2025, pp. 1-14

2502.08037 2025-02-13 cs.CL 57%

Franken-Adapter: Cross-Lingual Adaptation of LLMs by Embedding Surgery

Fan Jiang, Honglin Yu, Grace Chung, Trevor Cohn

机构 * The University of Melbourne（墨尔本大学）； Google（谷歌公司）

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments 33 pages

2502.05556 2025-02-11 cs.AI 57%

Knowledge is Power: Harnessing Large Language Models for Enhanced Cognitive Diagnosis

Zhiang Dong, Jingyuan Chen, Fei Wu

专题命中其他安全：alignment(abstract)；分类 cs.AI

2502.05458 2025-02-11 cs.CV cs.LG stat.ML 57%

Block Graph Neural Networks for tumor heterogeneity prediction

Marianne Abémgnigni Njifon, Tobias Weber, Viktor Bezborodov, Tyll Krueger, Dominic Schuhmacher

机构 * Institute for Mathematical Stochastics（数学随机学研究所）； University of Göttingen（哥廷根大学）； Tübingen AI Center（蒂宾根人工智能中心）； University of Tübingen（蒂宾根大学）； Wrocław University of Science and Technology（弗罗茨瓦夫理工大学）

专题命中其他安全：alignment(abstract)；分类 cs.LG

Comments 27 pages, 8 figures

2410.06176 2025-02-11 cs.CR cs.AI 57%

SC-Bench: A Large-Scale Dataset for Smart Contract Auditing

Shihao Xia, Mengting He, Linhai Song, Yiying Zhang

机构 * The Pennsylvania State University（宾夕法尼亚州立大学）； University of California, San Diego（加利福尼亚大学圣迭戈分校）

专题命中其他安全：safety(abstract)；分类 cs.AI

2408.05307 2025-02-11 cs.CE cs.LG 57%

Audio-visual cross-modality knowledge transfer for machine learning-based in-situ monitoring in laser additive manufacturing

Jiarui Xie, Mutahar Safdar, Lequn Chen, Seung Ki Moon, Yaoyao Fiona Zhao

专题命中其他安全：alignment(abstract)；分类 cs.LG

Comments 47 pages, 19 figures, 6 tables

2306.09996 2025-02-11 cs.CV cs.CL 57%

Investigating Prompting Techniques for Zero- and Few-Shot Visual Question Answering

Rabiul Awal, Le Zhang, Aishwarya Agrawal

专题命中其他安全：alignment(abstract)；分类 cs.CL

Comments Codes available at https://github.com/rabiulcste/vqazero

2502.04110 2025-02-07 cs.HC cs.AI 57%

Ancient Greek Technology: An Immersive Learning Use Case Described Using a Co-Intelligent Custom ChatGPT Assistant

Vlasis Kasapakis, Leonel Morgado

专题命中其他安全：alignment(abstract)；分类 cs.AI

Comments 5 pages, presented at the 2024 IEEE 3rd International Conference on Intelligent Reality (ICIR 2024), 6th of December, 2024

2502.04008 2025-02-07 cs.SE cs.AI 57%

Automating a Complete Software Test Process Using LLMs: An Automotive Case Study

Shuai Wang, Yinan Yu, Robert Feldt, Dhasarathy Parthasarathy

机构 * Chalmers University of Technology（查尔姆斯理工大学）； Volvo Group（沃尔沃集团）

专题命中其他安全：alignment(abstract)；分类 cs.AI

Comments Accepted by International Conference on Software Engineering (ICSE) 2025

1. 其他安全 8057 篇

Analyzing Factors Influencing Driver Willingness to Accept Advanced Driver Assistance Systems

An Efficient Sign Language Translation Using Spatial Configuration and Motion Dynamics with LLMs

Large Language Model as a Teacher for Zero-shot Tagging at Extreme Scales

Improving Consistency in Large Language Models through Chain of Guidance

LaVida Drive: Vision-Text Interaction VLM for Autonomous Driving with Token Selection, Recovery and Enhancement

Textual-to-Visual Iterative Self-Verification for Slide Generation

ChatVLA: Unified Multimodal Understanding and Robot Control with Vision-Language-Action Model

EAGER-LLM: Enhancing Large Language Models as Recommenders through Exogenous Behavior-Semantic Integration

Predicting Filter Medium Performances in Chamber Filter Presses with Digital Twins Using Neural Network Technologies

Multi-Faceted Studies on Data Poisoning can Advance LLM Development

An Open-Source Web-Based Tool for Evaluating Open-Source Large Language Models Leveraging Information Retrieval from Custom Documents

An Information-Theoretic Analysis of Thompson Sampling for Logistic Bandits

LLM4TS: Aligning Pre-Trained LLMs as Data-Efficient Time-Series Forecasters

Integration of Agentic AI with 6G Networks for Mission-Critical Applications: Use-case and Challenges

Prompting a Weighting Mechanism into LLM-as-a-Judge in Two-Step: A Case Study

Fake It Till You Make It: Using Synthetic Data and Domain Knowledge for Improved Text-Based Learning for LGE Detection

Balanced Multi-Factor In-Context Learning for Multilingual Large Language Models

Structure-aware Domain Knowledge Injection for Large Language Models

Grounded Knowledge-Enhanced Medical Vision-Language Pre-training for Chest X-Ray

Small Models, Big Impact: Efficient Corpus and Graph-Based Adaptation of Small Multilingual Language Models for Low-Resource Languages

Data2Concept2Text: An Explainable Multilingual Framework for Data Analysis Narration

On LLM-generated Logic Programs and their Inference Execution Methods

Franken-Adapter: Cross-Lingual Adaptation of LLMs by Embedding Surgery

Knowledge is Power: Harnessing Large Language Models for Enhanced Cognitive Diagnosis

Block Graph Neural Networks for tumor heterogeneity prediction

SC-Bench: A Large-Scale Dataset for Smart Contract Auditing

Audio-visual cross-modality knowledge transfer for machine learning-based in-situ monitoring in laser additive manufacturing

Investigating Prompting Techniques for Zero- and Few-Shot Visual Question Answering

Ancient Greek Technology: An Immersive Learning Use Case Described Using a Co-Intelligent Custom ChatGPT Assistant

Automating a Complete Software Test Process Using LLMs: An Automotive Case Study