{"ID":2861263,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.01623","arxiv_id":"2510.01623","title":"VLA-R1: Enhancing Reasoning in Vision-Language-Action Models","abstract":"Vision-Language-Action (VLA) models aim to unify perception, language understanding, and action generation, offering strong cross-task and cross-scene generalization with broad impact on embodied AI. However, current VLA models often lack explicit step-by-step reasoning, instead emitting final actions without considering affordance constraints or geometric relations. Their post-training pipelines also rarely reinforce reasoning quality, relying primarily on supervised fine-tuning with weak reward design. To address these challenges, we present VLA-R1, a reasoning-enhanced VLA that integrates Reinforcement Learning from Verifiable Rewards (RLVR) with Group Relative Policy Optimization (GRPO) to systematically optimize both reasoning and execution. Specifically, we design an RLVR-based post-training strategy with verifiable rewards for region alignment, trajectory consistency, and output formatting, thereby strengthening reasoning robustness and execution accuracy. Moreover, we develop VLA-CoT-13K, a high-quality dataset that provides chain-of-thought supervision explicitly aligned with affordance and trajectory annotations. Furthermore, extensive evaluations on in-domain, out-of-domain, simulation, and real-robot platforms demonstrate that VLA-R1 achieves superior generalization and real-world performance compared to prior VLA methods. We plan to release the model, code, and dataset following the publication of this work. Code: https://github.com/GigaAI-research/VLA-R1. Website: https://gigaai-research.github.io/VLA-R1.","short_abstract":"Vision-Language-Action (VLA) models aim to unify perception, language understanding, and action generation, offering strong cross-task and cross-scene generalization with broad impact on embodied AI. However, current VLA models often lack explicit step-by-step reasoning, instead emitting final actions without consideri...","url_abs":"https://arxiv.org/abs/2510.01623","url_pdf":"https://arxiv.org/pdf/2510.01623v1","authors":"[\"Angen Ye\",\"Zeyu Zhang\",\"Boyuan Wang\",\"Xiaofeng Wang\",\"Dapeng Zhang\",\"Zheng Zhu\"]","published":"2025-10-02T02:54:03Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.RO\"]","methods":"[\"Reinforcement Learning\"]","has_code":false,"code_links":[{"ID":608803,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2861263,"paper_url":"https://arxiv.org/abs/2510.01623","paper_title":"VLA-R1: Enhancing Reasoning in Vision-Language-Action Models","repo_url":"https://github.com/GigaAI-research/VLA-R1","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}