Glossary
What is SAC? Definition and meaning
An off-policy actor-critic algorithm that maximizes reward plus policy entropy, which encourages exploration and stability.
Definition
SAC: An off-policy actor-critic algorithm that maximizes reward plus policy entropy, which encourages exploration and stability.
In short: Soft actor-critic: off-policy and entropy-regularized.
When it is used
Continuous control with good sample efficiency, such as energy and process optimization.
Learn more
The full guide to SAC explains how it works, when to use it and how it compares with the alternatives, with examples. The glossary lists every term on the site from A to Z.
Previous term: S4 · Next term: Saga pattern