Vector Quantized Models for Planning

Sherjil Ozair · Yazhe Li · Ali Razavi · Ioannis Antonoglou · AƤron van den Oord · Oriol Vinyals

[ Abstract ] [ Livestream: Visit Reinforcement Learning and Planning 2 ] [ Paper ]
Tue 20 Jul 7:40 a.m. — 7:45 a.m. PDT

Recent developments in the field of model-based RL have proven successful in a range of environments, especially ones where planning is essential. However, such successes have been limited to deterministic fully-observed environments. We present a new approach that handles stochastic and partially-observable environments. Our key insight is to use discrete autoencoders to capture the multiple possible effects of an action in a stochastic environment. We use a stochastic variant of Monte Carlo tree search to plan over both the agent's actions and the discrete latent variables representing the environment's response. Our approach significantly outperforms an offline version of MuZero on a stochastic interpretation of chess where the opponent is considered part of the environment. We also show that our approach scales to DeepMind Lab, a first-person 3D environment with large visual observations and partial observability.

Chat is not available.