Institute of Trustworthy Embodied AI（可信具身人工智能研究院）；Fudan University（复旦大学）；Shanghai Innovation Institute（上海创新研究院）；Ant Group（蚂蚁集团）；Zhejiang University（浙江大学）；City University of Hong Kong（香港城市大学）
Ming Wen, Yuxuan Liu, Kun Yang, Yunhao Feng, Zhuoer Xu, Yuhao Sun, Shiwen Cui, Xiang Zheng, Yi Liu, Xingjun Ma, Yu-Gang Jiang
On-policy self-distillation (OPSD) has emerged as an efficient post-training paradigm by using a teacher conditioned on privileged information to provide dense token-level supervision. Prior work has shown that OPSD can collapse in verifiable reasoning tasks, while safety alignment differs in that it is guided by high-level constitutions rather than explicit target answers. However, pilot studies reveals that safety OPSD nonetheless suffers from severe collapse, where constitutional conditioning contracts the teacher distribution toward short and overly conservative responses and Reverse KL further amplifies this contraction into reduced expressiveness. We formalize this effect as geometric leakage under safety boundaries in a non-orthogonal semantic space, where safety pressure transfers into the expressiveness dimension. Based on this analysis, we propose Constitutional On-Policy Safety Distillation (COPSD), which first calibrates the teacher through a Cross-SFT cold-start and then performs constitution-conditioned on-policy distillation. Experiments across 3 multimodal large language models (MLLMs) on 12 safety and general benchmarks show that COPSD improves both safety and helpfulness over baselines while reducing the safety tax on general reasoning.
SafeSteer: Localized On-Policy Distillation for Efficient Safety Alignment
Beihang University（北航）；Beijing Institute of Technology（北京理工大学）；Beijing University of Posts and Telecommunications（北京邮电大学）；Peking University（北京大学）；Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；Beijing Academy of Artificial Intelligence（北京人工智能研究院）
Distillation Traps and Guards: A Calibration Knob for LLM Distillability
Generative AI Lab, College of Computing and Data Science（生成人工智能实验室，计算与数据科学学院）；Nanyang Technological University（南洋理工大学）；Systems Research Institute of the Polish Academy of Sciences（波兰科学院系统研究所）；AGH University of Krakow（克拉科夫AGH大学）；SAN University（SAN大学）
Backtracking When It Strays: Mitigating Dual Exposure Biases in LLM Reasoning Distillation
College of Computer Science and Technology, Jilin University（吉林大学计算机科学与技术学院）；Key Laboratory of Symbolic Computation and Knowledge Engineering, MoE, Jilin University（吉林大学符号计算与知识工程重点实验室）；Tongyi Lab, Alibaba Group（阿里集团通义实验室）；College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；School of Artificial Intelligence, Jilin University（吉林大学人工智能学院）
Alignment and Safety in Large Language Models: Safety Mechanisms, Training Paradigms, and Emerging Challenges
Department of Statistics, University of Georgia（统计学系，佐治亚大学）；School of Computing, University of Georgia（计算学院，佐治亚大学）；Department of Biostatistics, Boston University（生物统计学系，波士顿大学）；Department of Epidemiology & Biostatistics, University of Georgia（流行病学与生物统计学系，佐治亚大学）；School of Computer and Cyber Sciences, Augusta University（计算机与网络科学学院，奥古斯塔大学）；School of Electrical and Computer Engineering, University of Georgia（电气与计算机工程学院，佐治亚大学）；Department of Statistics & Data Science, University of Arizona（统计学与数据科学系，亚利桑那大学）；Kellogg School of Management, Northwestern University（凯洛格管理学院，西北大学）；Department of Statistics, Harvard University（统计学系，哈佛大学）；School of Computer Science, Carnegie Mellon University（计算机科学学院，卡内基梅隆大学）
Adversarial Prompt Distillation for Vision-Language Models
Shanghai Key Lab of Intell. Info. Processing, School of CS, Fudan University（上海智能信息处理实验室，计算机科学学院，复旦大学）；The Chinese University of Hong Kong, Shatin, Hong Kong（香港中文大学，沙田，香港）；The University of Hong Kong, Pokfulam, Hong Kong（香港大学，薄扶林，香港）
HarmAug: Effective Data Augmentation for Knowledge Distillation of Safety Guard Models
KAIST（韩国科学技术院）；Theori；Université de Montréal（蒙特利尔大学）；Mila – Québec AI Institute（米拉-魁北克人工智能研究所）；Technische Hochschule Nürnberg Georg Simon Ohm（纽伦堡应用技术大学）；CIFAR AI Chair（加拿大高级研究所人工智能主席项目）
Multi-Label Image Classification via Knowledge Distillation from Weakly-Supervised Detection
National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所模式识别国家重点实验室）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；CUHK-SenseTime Joint Lab, The Chinese University of Hong Kong（香港中文大学-商汤联合实验室）；SenseTime Research（商汤科技研究院）
