大模型对齐与安全 - arXivDaily 专题

2606.12429 2026-06-12 cs.CY cs.AI 新提交 81%

Muse Spark Safety & Preparedness Report

Muse Spark 安全与准备报告

Cristina Menghini, Peter Ney, Hamza Kwisaba, Zifan, Wang, Miles Turpin, Felix Binder, Jean-Christophe Testud, Aidan Boyd, Nathaniel Li, Ivan Evtimov, Klaudia Krawiecka, Arman Zharmagambetov, Jeremy Kritz, Alexander R. Fabbri, Daniel Song, Jinpeng Miao, Joonas Hjelt, Meghna Ramani, Leona Lan, Reza Aghajani, Joanna Bitton, Mahesh Pasupuleti, Devin Norder, Khalid El-Arini, Paridhi Singh, Vítor Albiero, Sahana CB, Rashnil Chaturvedi, Elahe Dabir, Edoardo Debenedetti, Jim Gust, Ziwen Han, Kat He, Sean Hendryx, Lifeng Jin, Polina Kirichenko, Sandra Lefdal, Kenneth Li, Asad Liaqat, Inna Lin, Despoina Magka, Neal Mangaokar, Ishita Mediratta, Zach Miller, Smitha Milli, Niloofar Mireshghallah, Saba Nazir, Hung Nguyen, Maximilian Nickel, Kelvin Niu, Kerem Oktar, Bhargavi Paranjape, Parth Pathak, Maya Pavlova, Emmanuel Ramirez, David Renardy, Candace Ross, Yasha Sheynin, Claudia Shi, Shivam Singhal, Evangelia Spiliopoulou, Rakshith Sharma Srinivasa, Jamelle Watson-Daniels, Spencer Whitman, Adina Williams, Chen Xing, Andy Zou, Tommy Ma, Siqi Deng, James Beldock, Prashant Ratanchandani, Kate Plawiak, Taesung Lee, Ryan Victory, Lindsay Hundley, Rachad Alao, Himaghna Bhattacharjee, Jianfeng Chi, Gary Frost, Pegah Ghahremani, Niki Howe, Yuheng Huang, Saeed Jahed, Hannah Korevaar, Trang Le, Zhe Liu, Jinghong Luo, Qin Lyu, Nina Mehrabi, Abraham Montilla, Chirag Nagpal, Cyrus Nikolaidis, Rajvardhan Oak, Manoj Ravi, Vidya Sarma, Aman Shankar, Alana Shine, Eric Michael Smith, Mariana Tandon, Michael Tontchev, Caoyu Wang, Zihan Wang, Corinne Wong, Zheng Wu, Hongyuan Zhan, Justin Zhao, Zexuan Zhong, Chengxu Zhuang, Tristan Goodman, Ayaz Minhas, Harrison Rudolph, Victoria Jeffries, Ingrid Dickinson, Alex Vaughan, Lauren Deason, Kamalika Chaudhuri, Julian Michael, Shengjia Zhao, Summer Yue

机构 * Muse Spark

专题命中安全训练：safety(title,abstract)；分类 cs.AI、cs.CY

AI总结 Meta 发布 Muse Spark 大语言模型，评估其在化学/生物、网络安全和失控风险等灾难性风险领域的安全性，通过多层缓解措施将风险降至可接受水平，并作为 Meta AI 的基础模型发布。

Comments 159 pages, 57 figures

2606.13621 2026-06-12 cs.AI cs.CR cs.GT cs.LG cs.MA 新提交 62%

Beyond Runtime Enforcement: Shield Synthesis as Defensibility Analysis for Adversarial Networks

超越运行时强制：作为对抗网络可防御性分析的盾牌合成

Achraf Hsain, Sultan Almuhammadi

机构 * Information and Computer Science Department, King Fahd University of Petroleum and Minerals（信息与计算机科学系，法赫德国王石油矿产大学）

专题命中安全训练：safety(abstract)；分类 cs.AI、cs.LG

AI总结提出将盾牌合成重新解释为设计时分析工具，通过约束双人安全博弈生成可防御性判定，并融合拓扑度量和强化学习行为形成可防御性指纹，揭示系统安全的结构性见解。

Comments 26 pages, 7 figures, 7 tables. Under review at JAIR. Code: https://github.com/AchrafHsain7/Bastion

2606.12640 2026-06-12 cs.LG cs.RO cs.SY eess.SY 新提交 57%

Individual Control Barrier Functions-Guided Diffusion Model for Safe Offline Multi-Agent Reinforcement Learning

个体控制障碍函数引导的扩散模型用于安全离线多智能体强化学习

Qingyun Guo, Junyi Shi, Jianuo Huang, Tianyu Shi

机构 * Department of Electrical Engineering and Automation, Aalto University（阿尔托大学电气工程与自动化系）； School of Computing and Data Science, Xiamen University Malaysia（厦门大学马来西亚分校计算与数据科学学院）； Department of Computer Science, University of Toronto（多伦多大学计算机科学系）

专题命中安全训练：safety(abstract)；分类 cs.LG

AI总结提出一种将神经个体控制障碍函数嵌入扩散模型的离线多智能体强化学习算法，通过逆动力学恢复控制策略，在保证奖励的同时显著提升轨迹生成的安全性。

Comments Accepted to the 23rd IFAC World Congress, 2026

1. 安全训练 3 篇

Muse Spark Safety & Preparedness Report

Beyond Runtime Enforcement: Shield Synthesis as Defensibility Analysis for Adversarial Networks

Individual Control Barrier Functions-Guided Diffusion Model for Safe Offline Multi-Agent Reinforcement Learning