Reinforcement Learning with Deep Energy-Based Policies

TL;DR

This paper introduces soft Q-learning, a novel reinforcement learning algorithm that learns energy-based policies for continuous states and actions, enabling improved exploration and skill transfer in robotic tasks.

Contribution

It presents a new method for learning expressive energy-based policies in continuous domains, combining maximum entropy policies with amortized Stein variational gradient descent.

Findings

Enhanced exploration capabilities demonstrated in robotic simulations

Successful transfer of skills across different tasks

Connection established between energy-based policies and actor-critic methods

Abstract

We propose a method for learning expressive energy-based policies for continuous states and actions, which has been feasible only in tabular domains before. We apply our method to learning maximum entropy policies, resulting into a new algorithm, called soft Q-learning, that expresses the optimal policy via a Boltzmann distribution. We use the recently proposed amortized Stein variational gradient descent to learn a stochastic sampling network that approximates samples from this distribution. The benefits of the proposed algorithm include improved exploration and compositionality that allows transferring skills between tasks, which we confirm in simulated experiments with swimming and walking robots. We also draw a connection to actor-critic methods, which can be viewed performing approximate inference on the corresponding energy-based model.