#视频-动作模型 (1)

Zero-WAM: In-Context World-Action Modeling from Human Videos for Open-Ended Task Generalization