Phoenix: Learning Scene Structure Representations for Generalizable World-Action Modelswith Yupeng Zheng, Chaoyue Li, Linbo Wang, Xiang Li +131 versionW_qjf8tvhdv1 · current