English

Extreme Q-Learning: MaxEnt RL without Entropy

Machine Learning 2023-03-02 v2 Artificial Intelligence Robotics

Abstract

Modern Deep Reinforcement Learning (RL) algorithms require estimates of the maximal Q-value, which are difficult to compute in continuous domains with an infinite number of possible actions. In this work, we introduce a new update rule for online and offline RL which directly models the maximal value using Extreme Value Theory (EVT), drawing inspiration from economics. By doing so, we avoid computing Q-values using out-of-distribution actions which is often a substantial source of error. Our key insight is to introduce an objective that directly estimates the optimal soft-value functions (LogSumExp) in the maximum entropy RL setting without needing to sample from a policy. Using EVT, we derive our \emph{Extreme Q-Learning} framework and consequently online and, for the first time, offline MaxEnt Q-learning algorithms, that do not explicitly require access to a policy or its entropy. Our method obtains consistently strong performance in the D4RL benchmark, outperforming prior works by \emph{10+ points} on the challenging Franka Kitchen tasks while offering moderate improvements over SAC and TD3 on online DM Control tasks. Visualizations and code can be found on our website at https://div99.github.io/XQL/.

Keywords

Cite

@article{arxiv.2301.02328,
  title  = {Extreme Q-Learning: MaxEnt RL without Entropy},
  author = {Divyansh Garg and Joey Hejna and Matthieu Geist and Stefano Ermon},
  journal= {arXiv preprint arXiv:2301.02328},
  year   = {2023}
}

Comments

ICLR 2023 Oral

R2 v1 2026-06-28T08:04:31.382Z