Unified Embodied VLM Reasoning with Robotic Action via Autoregressive Discretized Pre-training

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