GRAM: Generalization in Deep RL with a Robust Adaptation Module

TL;DR

This paper introduces GRAM, a framework for deep reinforcement learning that enhances generalization across both familiar and novel environment dynamics using a robust adaptation module, validated through simulation and hardware tests.

Contribution

The paper proposes a unified architecture with a robust adaptation module for dynamics generalization in deep RL, combining in-distribution adaptation and out-of-distribution robustness.

Findings

Strong generalization in simulation and hardware tests

Effective handling of both in-distribution and out-of-distribution scenarios

Improved robustness in quadruped robot locomotion

Abstract

The reliable deployment of deep reinforcement learning in real-world settings requires the ability to generalize across a variety of conditions, including both in-distribution scenarios seen during training as well as novel out-of-distribution scenarios. In this work, we present a framework for dynamics generalization in deep reinforcement learning that unifies these two distinct types of generalization within a single architecture. We introduce a robust adaptation module that provides a mechanism for identifying and reacting to both in-distribution and out-of-distribution environment dynamics, along with a joint training pipeline that combines the goals of in-distribution adaptation and out-of-distribution robustness. Our algorithm GRAM achieves strong generalization performance across in-distribution and out-of-distribution scenarios upon deployment, which we demonstrate through extensive simulation and hardware locomotion experiments on a quadruped robot.