IRL-VLA: Training an Vision-Language-Action Policy via Reward World Model

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