What to Ignore, What to React: Visually Robust RL Fine-Tuning of VLA Models

返回详情 VLA / Vision-Language-Action 每日论文卡