Self-Improving Vision-Language-Action Models with Data Generation via Residual RL
返回详情
VLA / Vision-Language-Action 每日论文卡