1 paper
Hanwei Liu, Huijia Zhang, Hanzhang Wang +6
Embodied video world models predict scene changes caused by robot actions, yet generated object responses may remain inconsistent with task intent and contact dynamics. Critical re…