CARE: Multi-Task Pretraining for Latent Continuous Action Representation in Robot Control

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