AI RESEARCH
Co-Evolving Latent Action World Models
arXiv CS.LG
•
ArXi:2510.26433v2 Announce Type: replace Adapting pretrained video generation models into controllable world models via latent actions is a promising step towards creating generalist world models. The dominant paradigm adopts a two-stage approach that trains latent action model (LAM) and the world model separately, resulting in redundant