Discrete Latent Space World Models

World models, which learn a compressed representation of a dynamic environment like, say, a video game, have delivered top results in reinforcement learning. A new method makes them much smaller.