English

Revisiting Discrete Soft Actor-Critic

Machine Learning 2024-11-21 v4 Artificial Intelligence

Abstract

We study the adaption of Soft Actor-Critic (SAC), which is considered as a state-of-the-art reinforcement learning (RL) algorithm, from continuous action space to discrete action space. We revisit vanilla discrete SAC and provide an in-depth understanding of its Q value underestimation and performance instability issues when applied to discrete settings. We thereby propose Stable Discrete SAC (SDSAC), an algorithm that leverages entropy-penalty and double average Q-learning with Q-clip to address these issues. Extensive experiments on typical benchmarks with discrete action space, including Atari games and a large-scale MOBA game, show the efficacy of our proposed method. Our code is at: https://github.com/coldsummerday/SD-SAC.git.

Keywords

Cite

@article{arxiv.2209.10081,
  title  = {Revisiting Discrete Soft Actor-Critic},
  author = {Haibin Zhou and Tong Wei and Zichuan Lin and junyou li and Junliang Xing and Yuanchun Shi and Li Shen and Chao Yu and Deheng Ye},
  journal= {arXiv preprint arXiv:2209.10081},
  year   = {2024}
}

Comments

Accepted by Transactions on Machine Learning Research (TMLR)

R2 v1 2026-06-28T01:47:04.097Z