1 paper
Manish Kumar Govind, Dominick Reilly, Smit Patel +2
Going beyond predicting robot actions, World Action Models (WAMs) can also generate future visual observations. We build on this generative capability to propose Recurrent Generati…