Reinforcement learning (RL) has emerged as a promising paradigm for inducing explicit reasoning behaviors in large language and vision-language models. However, reasoning-oriented RL post-training remains fundamentally challenging due to sparse trajectory-level rewards, leading to ambiguous credit assignment and severe exploration failures that can trap the policy in a ``learning cliff.'' Recent on-policy distillation methods introduce dense teacher supervision to stabilize optimization, but apply it uniformly across all generated trajectories. We argue that such uniform distillation is ill-suited for reasoning-intensive tasks, as low-quality on-policy trajectories often originate from early logical errors, and distillation under flawed contexts injects noisy and misaligned gradients. To address these challenges, we propose Knowledge-Enhanced Preference Optimization (KEPO), a unified post-training framework that integrates: (i) a quality-gated on-policy distillation objective that selectively applies dense teacher guidance only to high-quality trajectories, and (ii) a knowledge-enhanced exploration strategy that leverages hints learned from a teacher model to rejectively sample reward-positive on-policy trajectories for RL, thereby mitigating exploration collapse. Evaluated on a challenging medical visual question answering benchmark under single-source generalization, KEPO demonstrates improved training stability, more coherent reasoning behaviors, and superior out-of-distribution performance over reinforcement learning and on-policy distillation baselines.
MedAlign: A Synergistic Framework of Multimodal Preference Optimization and Federated Meta-Cognitive Reasoning
School of Automation, Guangdong University of Technology（广东工业大学自动化学院）；College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics（南京航空航天大学计算机科学与技术学院）；State Key Laboratory of Traditional Chinese Medicine Syndrome, The Second Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangdong Provincial Hospital of Chinese Medicine, Guangdong Provincial Academy of Chinese Medical Sciences（广东省中医药科学院中医证候重点实验室，广州中医药大学第二附属医院，广东省中医院，广东省中医药科学院）；College of Computing and Data Science, Nanyang Technological University（南洋理工大学计算机与数据科学学院）；Department of Electrical and Computer Engineering, Sungkyunkwan University（成均馆大学电子与计算机工程系）
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
Stanford University（斯坦福大学）；Emory University（埃默里大学）；University of California, Berkeley（加州大学伯克利分校）；University of Washington（华盛顿大学）；Massachusetts General Hospital（麻省总医院）；Harvard Medical School（哈佛医学院）
Bridging the Semantic-Utility Gap in Multimodal RAG via Generator-in-the-Loop Alignment
Elmore Family School of Electrical and Computer Engineering, Purdue University（普渡大学埃尔莫尔电气与计算机工程学院）；Yonsei University（延世大学）；University at Buffalo–SUNY（纽约州立大学布法罗分校）
MMUnlearner: Reformulating Multimodal Machine Unlearning in the Era of Multimodal Large Language Models
The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；The Hong Kong University of Science and Technology（香港科技大学）；Tongji University（同济大学）
Epistemic-aware Vision-Language Foundation Model for Fetal Ultrasound Interpretation
National Engineering Research Center for Multimedia Software, School of Computer Science, Wuhan University（国家多媒体软件工程技术研究中心，武汉大学计算机学院）；College of Computing and Data Science, Nanyang Technological University（computing and Data Science学院，南洋理工大学）
