大模型对齐与安全

Mirage or Method? How Model-Task Alignment Induces Divergent RL Conclusions

Haoze Wu, Cheng Wang, Wenshuo Zhao, Junxian He

机构 * Zhejiang University（浙江大学） ； National University of Singapore（国立新加坡大学） ； HKUST（香港科技大学） 纠错

专题命中 其他安全 ：alignment(title,abstract)；分类 cs.CL、cs.LG

Assessing AI-Generated Questions' Alignment with Cognitive Frameworks in Educational Assessment

Antoun Yaacoub, Jérôme Da-Rugna, Zainab Assaghir

专题命中 其他安全 ：alignment(title,abstract)；分类 cs.CL、cs.AI

Comments This paper was presented in the 17th Int. Conf. on Computer Science and Information Technology (ICCSIT 2024), Dubai, United Arab Emirates, 2024, Oct. 23-25. IT was published in the International Journal of Computer Theory and Engineering, vol. 17, no. 3, pp. 114-125, 2025

Journal ref International Journal of Computer Theory and Engineering, vol. 17, no. 3, pp. 114-125, 2025

Predicting NCAP Safety Ratings: An Analysis of Vehicle Characteristics and ADAS Features Using Machine Learning

Raunak Kunwar, Aera Kim LeBoulluec

专题命中 其他安全 ：safety(title,abstract)；分类 cs.LG

Comments 11 pages, 4 figures, Under review

BALM-TSF: Balanced Multimodal Alignment for LLM-Based Time Series Forecasting

Shiqiao Zhou, Holger Schöner, Huanbo Lyu, Edouard Fouché, Shuo Wang

机构 * University of Birmingham（伯明翰大学） 纠错

专题命中 其他安全 ：alignment(title,abstract)；分类 cs.AI

Beyond Pixels: Introducing Geometric-Semantic World Priors for Video-based Embodied Models via Spatio-temporal Alignment

Jinzhou Tang, Jusheng zhang, Sidi Liu, Waikit Xiu, Qinhan Lv, Xiying Li

专题命中 其他安全 ：alignment(title,abstract)；分类 cs.AI

Self-supervised large-scale kidney abnormality detection in drug safety assessment studies

Ivan Slootweg, Natalia P. García-De-La-Puente, Geert Litjens, Salma Dammak

机构 * Department of Pathology, Radboud University Medical Center（拉德堡德大学医学中心病理科部） ； Instituto Universitario de Investigación en Tecnología Centrada en el Ser Humano, Universitat Politècnica de València（人类中心技术大学研究机构，巴塞罗那理工大学） 纠错

专题命中 其他安全 ：safety(title,abstract)

Statutory Construction and Interpretation for Artificial Intelligence

Luxi He, Nimra Nadeem, Michel Liao, Howard Chen, Danqi Chen, Mariano-Florentino Cuéllar, Peter Henderson

机构 * Princeton University（普林斯顿大学） ； Carnegie Endowment for International Peace（国际和平研究所） 纠错

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

Extracting OPQRST in Electronic Health Records using Large Language Models with Reasoning

Zhimeng Luo, Abhibha Gupta, Adam Frisch, Daqing He

机构 * School of Computing and Information University of Pittsburgh（计算与信息学院 西弗吉尼亚大学） ； Department of Emergency Medicine University of Pittsburgh（急诊医学系 西弗吉尼亚大学） 纠错

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

Securing Radiation Detection Systems with an Efficient TinyML-Based IDS for Edge Devices

Einstein Rivas Pizarro, Wajiha Zaheer, Li Yang, Khalil El-Khatib, Glenn Harvel

机构 * Ontario Tech University（安大略技术大学） 纠错

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

Comments Preprint author original pre review. Accepted and Presented at NPIC & HMIT 2025. The official proceedings version is available in the ANS Digital Library

Journal ref Proc. NPIC HMIT 2025 pp. 651-661 (2025)

Target-Oriented Single Domain Generalization

Marzi Heidari, Yuhong Guo

机构 * School of Computer Science, Carleton University（计算机科学学院，卡尔顿大学） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.AI、cs.LG

Improving Significant Wave Height Prediction Using Chronos Models

Yilin Zhai, Hongyuan Shi, Chao Zhan, Qing Wang, Zaijin You, Nan Wang

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

Comments arXiv admin note: text overlap with arXiv:2403.07815 by other authors

Journal ref Ocean Engineering, Volume 341, Part 2, 1 December 2025, Article 122502

Empowering Large Language Model for Sequential Recommendation via Multimodal Embeddings and Semantic IDs

Yuhao Wang, Junwei Pan, Xinhang Li, Maolin Wang, Yuan Wang, Yue Liu, Dapeng Liu, Jie Jiang, Xiangyu Zhao

机构 * City University of Hong Kong（香港城市大学） ； Tencent Inc.（腾讯公司） ； Tsinghua University（清华大学） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.AI

Comments CIKM 2025 Full Research Paper

M3Ret: Unleashing Zero-shot Multimodal Medical Image Retrieval via Self-Supervision

Che Liu, Zheng Jiang, Chengyu Fang, Heng Guo, Yan-Jie Zhou, Jiaqi Qu, Le Lu, Minfeng Xu

机构 * DAMO Academy, Alibaba Group（达摩院，阿里巴巴集团） ； Imperial College London（帝国理工学院） ； Tsinghua University（清华大学） ； Hupan Lab（华潘实验室） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.LG

Comments Technical Report

Zero-shot Cross-lingual NER via Mitigating Language Difference: An Entity-aligned Translation Perspective

Zhihao Zhang, Sophia Yat Mei Lee, Dong Zhang, Shoushan Li, Guodong Zhou

机构 * School of Computer Science & Technology, NLP Lab, Soochow University, China（计算机科学与技术学院、自然语言处理实验室、苏州大学） ； Department of Chinese and Bilingual Studies, The Hong Kong Polytechnic University（中文与双语研究系、香港理工大学） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL

Comments EMNLP 2025

Will You Be Aware? Eye Tracking-Based Modeling of Situational Awareness in Augmented Reality

Zhehan Qu, Tianyi Hu, Christian Fronk, Maria Gorlatova

机构 * Department of Computer Science Duke University（计算机科学系 哥伦比亚大学） ； Department of Electrical and Computer Engineering Duke University（电气与计算机工程系 哥伦比亚大学） 纠错

专题命中 其他安全 ：safety(abstract)；分类 cs.LG

XAI-Driven Machine Learning System for Driving Style Recognition and Personalized Recommendations

Feriel Amel Sellal, Ahmed Ayoub Bellachia, Meryem Malak Dif, Enguerrand De Rautlin De La Roy, Mouhamed Amine Bouchiha, Yacine Ghamri-Doudane

机构 * L3i - La Rochelle University（拉罗什-拉科什大学L3i） 纠错

专题命中 其他安全 ：safety(abstract)；分类 cs.LG

Fusion to Enhance: Fusion Visual Encoder to Enhance Multimodal Language Model

Yifei She, Huangxuan Wu

机构 * Beijing University of Posts and Telecommunications（北京邮电大学） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.AI

GIER: Gap-Driven Self-Refinement for Large Language Models

Rinku Dewri

机构 * University of Denver（丹佛大学） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL

Generative AI for Industrial Contour Detection: A Language-Guided Vision System

Liang Gong, Tommy, Wang, Sara Chaker, Yanchen Dong, Fouad Bousetouane, Brenden Morton, Mark Mendez

机构 * The University of Chicago（芝加哥大学） ； FabTrack 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.AI

Comments 20 pages, 5 figures

SEAL: Structure and Element Aware Learning to Improve Long Structured Document Retrieval

Xinhao Huang, Zhibo Ren, Yipeng Yu, Ying Zhou, Zulong Chen, Zeyi Wen

机构 * HKUST (Guangzhou)（香港科技大学（广州）） ； HKUST（香港科技大学） ； Alibaba Group（阿里巴巴集团） ； Zhejiang Lab（浙江实验室） 纠错

专题命中 其他安全 ：alignment(abstract)；分类 cs.LG

Comments Accepted at EMNLP 2025 Main Conference

OpenMulti: Open-Vocabulary Instance-Level Multi-Agent Distributed Implicit Mapping

Jianyu Dou, Yinan Deng, Jiahui Wang, Xingsi Tang, Yi Yang, Yufeng Yue

机构 * School of Automation, Beijing Institute of Technology（自动化学院，北京理工大学） 纠错

专题命中 其他安全 ：alignment(abstract)

Comments Accepted to IEEE Robotics and Automation Letters. Project website: https://openmulti666.github.io/

Beyond Gloss: A Hand-Centric Framework for Gloss-Free Sign Language Translation

Sobhan Asasi, Mohamed Ilyas Lakhal, Ozge Mercanoglu Sincan, Richard Bowden

机构 * Center for Vision, Speech and Signal Processing (CVSSP) University of Surrey（视觉、语音和信号处理中心（CVSSP）大学学院） 纠错

专题命中 其他安全 ：alignment(abstract)

Comments Accepted at BMVC 2025

NetRoller: Interfacing General and Specialized Models for End-to-End Autonomous Driving

Ren Xin, Hongji Liu, Xiaodong Mei, Wenru Liu, Maosheng Ye, Zhili Chen, Jun Ma

机构 * The Hong Kong University of Science and Technology（香港科学与技术大学） 纠错

专题命中 其他安全 ：safety(abstract)

Comments This work has been submitted to the IEEE for possible publication

HERMES: Human-to-Robot Embodied Learning from Multi-Source Motion Data for Mobile Dexterous Manipulation

Zhecheng Yuan, Tianming Wei, Langzhe Gu, Pu Hua, Tianhai Liang, Yuanpei Chen, Huazhe Xu

机构 * Tsinghua University（清华大学） ； Shanghai Qi Zhi Institute（上海启智研究院） ； Peking University（北京大学） 纠错

专题命中 其他安全 ：alignment(abstract)

Leveraging Modality Tags for Enhanced Cross-Modal Video Retrieval

Adriano Fragomeni, Dima Damen, Michael Wray

机构 * School of Computer Science University of Bristol（计算机科学学院英国布里斯托尔大学） 纠错

专题命中 其他安全 ：alignment(abstract)

Comments Accepted at BMVC 2025