The Nature of Temporal Difference Errors in Multi-step Distributional Reinforcement LearningDownload PDF

Published: 31 Oct 2022, Last Modified: 11 Oct 2022NeurIPS 2022 AcceptReaders: Everyone
Keywords: Distributional Reinforcement Learning, Off-policy Learning
TL;DR: The first theoretical guarantees on multi-step distributional RL algorithms
Abstract: We study the multi-step off-policy learning approach to distributional RL. Despite the apparent similarity between value-based RL and distributional RL, our study reveals intriguing and fundamental differences between the two cases in the multi-step setting. We identify a novel notion of path-dependent distributional TD error, which is indispensable for principled multi-step distributional RL. The distinction from the value-based case bears important implications on concepts such as backward-view algorithms. Our work provides the first theoretical guarantees on multi-step off-policy distributional RL algorithms, including results that apply to the small number of existing approaches to multi-step distributional RL. In addition, we derive a novel algorithm, Quantile Regression-Retrace, which leads to a deep RL agent QR-DQN-Retrace that shows empirical improvements over QR-DQN on the Atari-57 benchmark. Collectively, we shed light on how unique challenges in multi-step distributional RL can be addressed both in theory and practice.
Supplementary Material: pdf
15 Replies