10 papers
Semifactual Credit-Augmented Policy Optimization
Junshu Pan, Zhizhang Fu, Shulin Huang +5
Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions remain sensitive to task-irrele…
RL Starts before RL: On Policy Distillation for Better Reinforcement Learning
Shuai Dong, Yongfu Zhu, Yuqi Xu +26
Reinforcement learning (RL) improves reasoning, but its performance depends on the policy from which training begins. We study on-policy distillation (OPD) as a preparation stage f…
SafeRI: Recognition and Intervention for Token-Level Safety Intervention in Large Vision Language Models
Caoyuan Ma, Tian Gu, Wenpu Liu +11
Existing safety alignment methods for vision-language models usually modify the model behavior globally: once the safety parameters are trained or loaded, they participate in both…
SafeCap: Improving LVLM Safety with Image Captioning Reinforcement Learning
Caoyuan Ma, Wenpu Liu, Weichu Xie +12
Large vision-language models (LVLMs) remain vulnerable to jailbreak attacks that exploit visual inputs to bypass safety alignment inherited from their language backbones. We propos…
InfoMem: Training Long-Context Memory Agents with Answer-Conditioned Information Gain
Tiancheng Han, Yong Li, Wuzhou Yu +2
Long-context tasks require LLMs to identify and preserve answer-relevant information from large contexts. Chunk-wise memory agents address this issue by sequentially reading docume…
Right Makes Might: Aligning Verified Hidden States Empowers RL Reasoning
Ziyue Wang, Aomufei Yuan, Yongfu Zhu +10
Reinforcement Learning from Verifiable Rewards (RLVR) has become the dominant approach for improving mathematical reasoning in large language models, yet current methods reduce eac…
Think-Reflect-Revise: A Policy-Guided Reflective Framework for Safety Alignment in Large Vision Language Models
Fenghua Weng, Chaochao Lu, Xia Hu +2
As multimodal reasoning improves the overall capabilities of Large Vision Language Models (LVLMs), recent studies have begun to explore safety-oriented reasoning, aiming to enhance…
COSMO-RL: Towards Trustworthy LMRMs via Joint Safety and Stability
Yizhuo Ding, Mingkang Chen, Qiuhua Liu +7
Large Multimodal Reasoning Models (LMRMs) are moving into real applications, where they must be both useful and safe. Safety is especially challenging in multimodal settings: image…
UniPruning: Unifying Local Metric and Global Feedback for Scalable Sparse LLMs
Yizhuo Ding, Wanying Qu, Jiawei Geng +2
Large Language Models (LLMs) achieve strong performance across diverse tasks but face prohibitive computational and memory costs. Pruning offers a promising path by inducing sparsi…
VTPerception-R1: Enhancing Multimodal Reasoning via Explicit Visual and Textual Perceptual Grounding
Yizhuo Ding, Mingkang Chen, Zhibang Feng +4
Multimodal large language models (MLLMs) often struggle to ground reasoning in perceptual evidence. We present a systematic study of perception strategies-explicit, implicit, visua…