Ego2World: Compiling Egocentric Cooking Videos into Executable Worlds for Belief-State Planning

ArXi:2605.13335v1 Announce Type: new Embodied agents in household environments must plan under partial observation: they need to remember objects, track state changes, and recover when actions fail. Existing benchmarks only partially test this ability. Egocentric video datasets capture realistic human activities but remain passive, while interactive simulators execution but rely on synthetic scenes and hand-crafted dynamics,