Value Function Decomposition for Iterative Design of Reinforcement Learning AgentsDownload PDF

Published: 31 Oct 2022, 18:00, Last Modified: 03 Jan 2023, 19:21NeurIPS 2022 AcceptReaders: Everyone
Keywords: reinforcement learning, explainable AI, machine learning, decision making, deep learning
TL;DR: Providing multiple examples and analyses, we demonstrate how to apply value decomposition and new metrics to actor critic algorithms, such as SAC, that allow common reinforcement learning problems to be diagnosed and resolved.
Abstract: Designing reinforcement learning (RL) agents is typically a difficult process that requires numerous design iterations. Learning can fail for a multitude of reasons and standard RL methods provide too few tools to provide insight into the exact cause. In this paper, we show how to integrate \textit{value decomposition} into a broad class of actor-critic algorithms and use it to assist in the iterative agent-design process. Value decomposition separates a reward function into distinct components and learns value estimates for each. These value estimates provide insight into an agent's learning and decision-making process and enable new training methods to mitigate common problems. As a demonstration, we introduce SAC-D, a variant of soft actor-critic (SAC) adapted for value decomposition. SAC-D maintains similar performance to SAC, while learning a larger set of value predictions. We also introduce decomposition-based tools that exploit this information, including a new reward \textit{influence} metric, which measures each reward component's effect on agent decision-making. Using these tools, we provide several demonstrations of decomposition's use in identifying and addressing problems in the design of both environments and agents. Value decomposition is broadly applicable and easy to incorporate into existing algorithms and workflows, making it a powerful tool in an RL practitioner's toolbox.
Supplementary Material: zip
14 Replies