{"ID":5438866,"CreatedAt":"2026-07-01T01:17:58.482524686Z","UpdatedAt":"2026-07-03T12:44:19.017960396Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.31599","arxiv_id":"2606.31599","title":"Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning","abstract":"Vision-language models (VLMs) combining reinforcement learning (RL) ignite remarkable progress in multimodal reasoning, yet still struggle with medical images, which typically exhibit extremely sparse visual evidence to inform clinical decision-making. We recognize that pruning visual tokens outside the grounding region greatly enhances medical reasoning. However, a united RL framework for active visual token pruning (VTP) and medical multimodal reasoning remains unestablished. Here, we propose a dual-stream RL framework, ViToS, to fulfill token pruning and question answering. ViToS trains one policy model with two task branches, where one focuses on grounding while the other conducts token-sparse reasoning after VTP. Furthermore, we solve the coupled policy learning problem by introducing the cross-feedback sequential optimization, avoiding gradient conflict and facilitating convergence of the shared policy model. Evaluated on seven medical benchmarks, our method reduces visual tokens to 77% of the original sequence length while achieving a 108.27% relative performance on Lingshu-7B and 104.16% relative performance on HuatuoGPT-Vision-7B. Overall, ViToS delivers superior performance and inference speedup, establishing an efficient paradigm for medical multimodal reasoning.","short_abstract":"Vision-language models (VLMs) combining reinforcement learning (RL) ignite remarkable progress in multimodal reasoning, yet still struggle with medical images, which typically exhibit extremely sparse visual evidence to inform clinical decision-making. We recognize that pruning visual tokens outside the grounding regio...","url_abs":"https://arxiv.org/abs/2606.31599","url_pdf":"https://arxiv.org/pdf/2606.31599v1","authors":"[\"Kaitao Chen\",\"Weiqian Zhao\",\"Jiamin Wu\",\"Qihao Zheng\",\"Shangquan Sun\",\"Chunfeng Song\",\"Xiaosong Wang\",\"Mu Zhou\",\"Mianxin Liu\"]","published":"2026-06-30T12:47:30Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[\"Reinforcement Learning\",\"Language Model\"]","has_code":false}