Utilizing Prior Solutions for Reward Shaping and Composition in Entropy-Regularized Reinforcement Learning

TL;DR

This paper introduces a unified framework for reward shaping and task composition in entropy-regularized reinforcement learning, enabling faster learning by leveraging prior solutions and task structures.

Contribution

It derives exact relations for optimal soft value functions, facilitating reward shaping and task composition in entropy-regularized RL, with validated experimental improvements.

Findings

Reward shaping accelerates learning in entropy-regularized RL.

Task composition enables solving new problems by combining prior solutions.

Theoretical relations improve understanding of value functions in complex RL settings.

Abstract

In reinforcement learning (RL), the ability to utilize prior knowledge from previously solved tasks can allow agents to quickly solve new problems. In some cases, these new problems may be approximately solved by composing the solutions of previously solved primitive tasks (task composition). Otherwise, prior knowledge can be used to adjust the reward function for a new problem, in a way that leaves the optimal policy unchanged but enables quicker learning (reward shaping). In this work, we develop a general framework for reward shaping and task composition in entropy-regularized RL. To do so, we derive an exact relation connecting the optimal soft value functions for two entropy-regularized RL problems with different reward functions and dynamics. We show how the derived relation leads to a general result for reward shaping in entropy-regularized RL. We then generalize this approach to derive an exact relation connecting optimal value functions for the composition of multiple tasks in entropy-regularized RL. We validate these theoretical contributions with experiments showing that reward shaping and task composition lead to faster learning in various settings.