On Learning Intrinsic Rewards for Policy Gradient Methods

TL;DR

This paper introduces a new algorithm for learning intrinsic rewards tailored for policy-gradient reinforcement learning agents, demonstrating improved performance in Atari and Mujoco environments over extrinsic rewards alone.

Contribution

It develops a novel method to learn intrinsic rewards for policy-gradient agents, extending the Optimal Rewards Framework to learning agents.

Findings

Improved performance in most Atari and Mujoco tasks.

Intrinsic rewards enhance learning efficiency.

Baseline agents with only extrinsic rewards performed less effectively.

Abstract

In many sequential decision making tasks, it is challenging to design reward functions that help an RL agent efficiently learn behavior that is considered good by the agent designer. A number of different formulations of the reward-design problem, or close variants thereof, have been proposed in the literature. In this paper we build on the Optimal Rewards Framework of Singh et.al. that defines the optimal intrinsic reward function as one that when used by an RL agent achieves behavior that optimizes the task-specifying or extrinsic reward function. Previous work in this framework has shown how good intrinsic reward functions can be learned for lookahead search based planning agents. Whether it is possible to learn intrinsic reward functions for learning agents remains an open problem. In this paper we derive a novel algorithm for learning intrinsic rewards for policy-gradient based learning agents. We compare the performance of an augmented agent that uses our algorithm to provide additive intrinsic rewards to an A2C-based policy learner (for Atari games) and a PPO-based policy learner (for Mujoco domains) with a baseline agent that uses the same policy learners but with only extrinsic rewards. Our results show improved performance on most but not all of the domains.