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