大模型对齐与安全

Trustworthy AI Psychotherapy: Multi-Agent LLM Workflow for Counseling and Explainable Mental Disorder Diagnosis

Mithat Can Ozgun, Jiahuan Pei, Koen Hindriks, Lucia Donatelli, Qingzhi Liu, Junxiao Wang

机构 * Vrije Universiteit Amsterdam（阿姆斯特丹自由大学） ； Wageningen University and Research（瓦赫宁根大学和研究中心） ； Guangzhou University（广州大学） 纠错

专题命中 安全评测 ：trustworthy(title,abstract)；分类 cs.AI

Comments This paper has been accepted by CIKM 2025 as a full paper

Adversarial Generation and Collaborative Evolution of Safety-Critical Scenarios for Autonomous Vehicles

Jiangfan Liu, Yongkang Guo, Fangzhi Zhong, Tianyuan Zhang, Zonglei Jing, Siyuan Liang, Jiakai Wang, Mingchuan Zhang, Aishan Liu, Xianglong Liu

机构 * Beihang University（北京航空航天大学） ； Nanyang Technological University（南洋理工大学） ； Zhongguancun Laboratory（中关村实验室） ； Henan University of Science and Technology（河南科技大学） 纠错

专题命中 安全评测 ：safety(title,abstract)

RADAR: A Reasoning-Guided Attribution Framework for Explainable Visual Data Analysis

Anku Rani, Aparna Garimella, Apoorv Saxena, Balaji Vasan Srinivasan, Paul Pu Liang

专题命中 安全评测 ：alignment(abstract)；trustworthy(abstract)；分类 cs.AI

ZPD-SCA: Unveiling the Blind Spots of LLMs in Assessing Students' Cognitive Abilities

Wenhan Dong, Zhen Sun, Yuemeng Zhao, Zifan Peng, Jun Wu, Jingyi Zheng, Yule Liu, Xinlei He, Yu Wang, Ruiming Wang, Xinyi Huang, Lei Mo

机构 * School of Psychology, South China Normal University（南方科技大学心理学院） ； Information Hub, Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）信息中心） ； School of AI, Guangzhou University（广州大学人工智能学院） ； College of Cyber Security, Jinan University（济南大学网络安全学院） 纠错

专题命中 安全评测 ：alignment(abstract)；分类 cs.CL、cs.AI、cs.CY

Interpretable Early Failure Detection via Machine Learning and Trace Checking-based Monitoring

Andrea Brunello, Luca Geatti, Angelo Montanari, Nicola Saccomanno

机构 * University of Udine, Italy（乌迪内大学） 纠错

专题命中 安全评测 ：safety(abstract)；分类 cs.AI、cs.LG

Comments Full version of the paper accepted for publication at the 28th European Conference on Artificial Intelligence (ECAI 2025)

Robustness Feature Adapter for Efficient Adversarial Training

Quanwei Wu, Jun Guo, Wei Wang, Yi Wang

机构 * Dongguan University of Technology（东莞科技学院） ； The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州）） 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments The paper has been accepted for presentation at ECAI 2025

LETToT: Label-Free Evaluation of Large Language Models On Tourism Using Expert Tree-of-Thought

Ruiyan Qi, Congding Wen, Weibo Zhou, Jiwei Li, Shangsong Liang, Lingbo Li

专题命中 安全评测 ：alignment(abstract)；分类 cs.CL、cs.AI

Explainable Prediction of the Mechanical Properties of Composites with CNNs

Varun Raaghav, Dimitrios Bikos, Antonio Rago, Francesca Toni, Maria Charalambides

机构 * Department of Mechanical Engineering, Imperial College London, UK（帝国理工学院机械工程系） ； Department of Computing, Imperial College London, UK（帝国理工学院计算系） ； Department of Informatics, King's College London, UK（伦敦国王学院信息学系） 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 9 pages, 6 figures. Accepted for publication at The 14th Conference on Prestigious Applications of Intelligent Systems (PAIS-2025)

The AI Model Risk Catalog: What Developers and Researchers Miss About Real-World AI Harms

Pooja S. B. Rao, Sanja Šćepanović, Dinesh Babu Jayagopi, Mauro Cherubini, Daniele Quercia

专题命中 安全评测 ：safety(abstract)；分类 cs.AI、cs.CY

Comments Accepted to AIES 2025

Situational Awareness as the Imperative Capability for Disaster Resilience in the Era of Complex Hazards and Artificial Intelligence

Hongrak Pak, Ali Mostafavi

机构 * organization= UrbanResilience.AI Lab, Zachry Department of Civil ； Environmental Engineering, Texas A\&M University , city= College Station , postcode= 77843 , state= TX , country= USA 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.AI、cs.CY

Hierarchical Vision-Language Learning for Medical Out-of-Distribution Detection

Runhe Lai, Xinhua Lu, Kanghao Chen, Qichao Chen, Wei-Shi Zheng, Ruixuan Wang

机构 * Peng Cheng Laboratory（鹏城实验室） ； Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）） ； University of Nottingham Malaysia（诺丁汉大学（马来西亚）） ； Key Laboratory of Machine Intelligence and Advanced Computing, MOE（机器智能与高级计算重点实验室） 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.AI

Comments 10 pages, 2 figures, Accepted by MICCAI2025

FRAME : Comprehensive Risk Assessment Framework for Adversarial Machine Learning Threats

Avishag Shapira, Simon Shigol, Asaf Shabtai

机构 * Ben-Gurion University of the Negev（贝内杰尔大学） 纠错

专题命中 安全评测 ：alignment(abstract)；分类 cs.LG

UI-Level Evaluation of ALLaM 34B: Measuring an Arabic-Centric LLM via HUMAIN Chat

Omer Nacar

机构 * NAMAA Community（NAMAA社区） ； Riyadh - KSA（利雅得-科威特） 纠错

专题命中 安全评测 ：safety(abstract)；分类 cs.CL

When Developer Aid Becomes Security Debt: A Systematic Analysis of Insecure Behaviors in LLM Coding Agents

Matous Kozak, Roshanak Zilouchian Moghaddam, Siva Sivaraman

机构 * Microsoft（微软公司） ； Czech Technical University in Prague（捷克技术大学） 纠错

专题命中 安全评测 ：safety(abstract)；分类 cs.AI

Comments 15 pages

Hidden in Plain Sight: Reasoning in Underspecified and Misspecified Scenarios for Multimodal LLMs

Qianqi Yan, Hongquan Li, Shan Jiang, Yang Zhao, Xinze Guan, Ching-Chen Kuo, Xin Eric Wang

机构 * University of California, Santa Cruz（加州大学圣克鲁兹分校） ； eBay 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.AI

AutoMisty: A Multi-Agent LLM Framework for Automated Code Generation in the Misty Social Robot

Xiao Wang, Lu Dong, Sahana Rangasrinivasan, Ifeoma Nwogu, Srirangaraj Setlur, Venugopal Govindaraju

机构 * State University of New York at Buffalo（纽约州立大学布法罗分校） ； Department of Computer Science and Engineering, Amrita School of Computing, Amrita Vishwa Vidyapeetham（计算机科学与工程系，阿米特拉学校 computing，阿米特拉世界大学） 纠错

专题命中 安全评测 ：alignment(abstract)；分类 cs.AI

Comments Accepted by IROS 2025

Adversarial Robustness in Two-Stage Learning-to-Defer: Algorithms and Guarantees

Yannis Montreuil, Axel Carlier, Lai Xing Ng, Wei Tsang Ooi

机构 * School of Computing, National University of Singapore, Singapore（新加坡国立大学计算机学院） ； IRIT, Université de Toulouse, CNRS, Toulouse INP, Toulouse, France（图卢兹大学IRIT研究所、法国国家科学研究中心、图卢兹INP） ； Institute for Infocomm Research, Agency for Science, Technology and Research, Singapore（新加坡科技研究局信息与通信研究所） 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.LG

Comments Accepted at the 42nd International Conference on Machine Learning (ICML 2025)

Hyperbolic Graph Neural Networks: A Review of Methods and Applications

Menglin Yang, Min Zhou, Tong Zhang, Jiahong Liu, Zhihao Li, Lujia Pan, Hui Xiong, Irwin King

机构 * Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）） ； Huawei Technologies Co., Ltd.（华为技术有限公司） ； The Chinese University of Hong Kong（香港中文大学） ； Zhejiang University（浙江大学） 纠错

专题命中 安全评测 ：trustworthy(abstract)；分类 cs.LG

Comments The latest draft was circulated under the title "Hyperbolic Graph Learning: A Comprehensive Review." The present arXiv version retains the original title, "Hyperbolic Graph Neural Networks: A Review of Methods and Applications," for consistency, while incorporating substantial revisions and extensions

Traveling Across Languages: Benchmarking Cross-Lingual Consistency in Multimodal LLMs

Hao Wang, Pinzhi Huang, Jihan Yang, Saining Xie, Daisuke Kawahara

专题命中 安全评测 ：分类 cs.CL、cs.AI、cs.LG；prompt injection(comments)

Comments The first version of this paper mistakenly included a prompt injection phrase, which was inappropriate and unprofessional. Although we corrected the version on arXiv and withdrew from the conference, my co-authors and university strongly request a full withdrawal. Given the situation, I no longer have the authority to manage this paper, and withdrawing it from arXiv is the most responsible action

Enhanced Drift-Aware Computer Vision Architecture for Autonomous Driving

Md Shahi Amran Hossain, Abu Shad Ahammed, Sayeri Mukherjee, Roman Obermaisser

机构 * Chair of Embedded Systems University of Siegen（嵌入系统教授会 University of Siegen） 纠错

专题命中 安全评测 ：safety(abstract)

SEER-VAR: Semantic Egocentric Environment Reasoner for Vehicle Augmented Reality

Yuzhi Lai, Shenghai Yuan, Peizheng Li, Jun Lou, Andreas Zell

专题命中 安全评测 ：alignment(abstract)

From Scores to Skills: A Cognitive Diagnosis Framework for Evaluating Financial Large Language Models

Ziyan Kuang, Feiyu Zhu, Maowei Jiang, Yanzhao Lai, Zelin Wang, Zhitong Wang, Meikang Qiu, Jiajia Huang, Min Peng, Qianqian Xie, Sophia Ananiadou

专题命中 安全评测 ：trustworthy(abstract)

QA-VLM: Providing human-interpretable quality assessment for wire-feed laser additive manufacturing parts with Vision Language Models

Qiaojie Zheng, Jiucai Zhang, Joy Gockel, Michael B. Wakin, Craig Brice, Xiaoli Zhang

机构 * Mechanical Engineering, Colorado School of Mines（机械工程，科罗拉多矿业学院） ； Electrical Engineering, Colorado School of Mines（电气工程，科罗拉多矿业学院） 纠错

专题命中 安全评测 ：trustworthy(abstract)

Requirements-Driven Automated Software Testing: A Systematic Review

Fanyu Wang, Chetan Arora, Chakkrit Tantithamthavorn, Kaicheng Huang, Aldeida Aleti

专题命中 安全评测 ：alignment(abstract)

Comments Accepted by TOSEM