AI Dynamics

Global AI News Aggregator

About

Contrastive World Models Learn Without Reconstructing Every Pixel

"Contrastive World Models" If world model has to reconstruct every pixel, it'll pretty much waste most of its capacity modeling irrelevant background noise. So this paper removes Google DeepMind's Dreamer pixel decoder and instead trains the latent state to identify features of

→ View original post on X — @askalphaxiv