1 paper
Xiaodi Huang, Ziyi Ding, Jingtian Wan +6
LeWM is a lightweight visual world model that learns latent dynamics end-to-end from pixels and ranks candidate action sequences by the distance between their predicted endpoints a…