DSAC: Distributional Soft Actor-Critic for Risk-Sensitive Reinforcement Learning
Abstract
We present Distributional Soft Actor-Critic (DSAC), a distributional reinforcement learning (RL) algorithm that combines the strengths of distributional information of accumulated rewards and entropy-driven exploration from Soft Actor-Critic (SAC) algorithm. DSAC models the randomness in both action and rewards, surpassing baseline performances on various continuous control tasks. Unlike standard approaches that solely maximize expected rewards, we propose a unified framework for risk-sensitive learning, one that optimizes the risk-related objective while balancing entropy to encourage exploration. Extensive experiments demonstrate DSAC’s effectiveness in enhancing agent performances for both risk-neutral and risk-sensitive control tasks.
Cite
Text
Ma et al. "DSAC: Distributional Soft Actor-Critic for Risk-Sensitive Reinforcement Learning." Journal of Artificial Intelligence Research, 2025. doi:10.1613/JAIR.1.17526Markdown
[Ma et al. "DSAC: Distributional Soft Actor-Critic for Risk-Sensitive Reinforcement Learning." Journal of Artificial Intelligence Research, 2025.](https://mlanthology.org/jair/2025/ma2025jair-dsac/) doi:10.1613/JAIR.1.17526BibTeX
@article{ma2025jair-dsac,
title = {{DSAC: Distributional Soft Actor-Critic for Risk-Sensitive Reinforcement Learning}},
author = {Ma, Xiaoteng and Chen, Junyao and Xia, Li and Yang, Jun and Zhao, Qianchuan and Zhou, Zhengyuan},
journal = {Journal of Artificial Intelligence Research},
year = {2025},
doi = {10.1613/JAIR.1.17526},
volume = {83},
url = {https://mlanthology.org/jair/2025/ma2025jair-dsac/}
}