Learning to Move Before Learning to Do: Task-Agnostic pretraining for VLAs figure
AlphaXiv 中文概览(可滚动查看)