1 paper
AgiBot Research Team, Renhang Liu, Wenzhi Zhao +42
World-action models (WAM) predict future states to guide robot actions, enabling learning from both action-free video and action-labeled interaction. Most inherit pretrained video…