Learned world models summarize an agent's experience to facilitate learning complex behaviors. While learning world models from high-dimensional sensory inputs is becoming feasible through deep learning, there are many potential ways for deriving behaviors from them. We present Dreamer, a reinforcement learning agent that solves long-horizon tasks from images purely by latent imagination. We efficiently learn behaviors by propagating analytic gradients of learned state values back through trajectories imagined in the compact state space of a learned world model. On 20 challenging visual control tasks, Dreamer exceeds existing approaches in data-efficiency, computation time, and final performance.
@article{arxiv.1912.01603,
title = {Dream to Control: Learning Behaviors by Latent Imagination},
author = {Danijar Hafner and Timothy Lillicrap and Jimmy Ba and Mohammad Norouzi},
journal= {arXiv preprint arXiv:1912.01603},
year = {2020}
}