Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor

TL;DR

The paper introduces the soft actor-critic algorithm, an off-policy deep reinforcement learning method based on maximum entropy principles, which improves sample efficiency, stability, and performance on continuous control tasks.

Contribution

It presents a novel stochastic actor-critic algorithm that combines off-policy updates with maximum entropy reinforcement learning, achieving state-of-the-art results and enhanced stability.

Findings

Outperforms prior methods on continuous control benchmarks

Achieves high stability across different random seeds

Demonstrates improved sample efficiency and convergence

Abstract

Model-free deep reinforcement learning (RL) algorithms have been demonstrated on a range of challenging decision making and control tasks. However, these methods typically suffer from two major challenges: very high sample complexity and brittle convergence properties, which necessitate meticulous hyperparameter tuning. Both of these challenges severely limit the applicability of such methods to complex, real-world domains. In this paper, we propose soft actor-critic, an off-policy actor-critic deep RL algorithm based on the maximum entropy reinforcement learning framework. In this framework, the actor aims to maximize expected reward while also maximizing entropy. That is, to succeed at the task while acting as randomly as possible. Prior deep RL methods based on this framework have been formulated as Q-learning methods. By combining off-policy updates with a stable stochastic actor-critic formulation, our method achieves state-of-the-art performance on a range of continuous control benchmark tasks, outperforming prior on-policy and off-policy methods. Furthermore, we demonstrate that, in contrast to other off-policy algorithms, our approach is very stable, achieving very similar performance across different random seeds.