EUVIMEDEUVIMED entdecken
Crossref

Improving Stochastic Action-Constrained Reinforcement Learning via Truncated Distributions

Roland Stolz, Michael Eichelbeck, Matthias Althoff

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (30) · 25617-25626

DOI: 10.1609/aaai.v40i30.39758 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Roland Stolz und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In reinforcement learning (RL), it is often advantageous to consider additional constraints on the action space to ensure safety or action relevance. Existing work on such action-constrained RL faces challenges regarding effective policy updates, computational efficiency, and predictable runtime. Recent work proposes to use truncated normal distributions for stochastic policy gradient methods. However, the computation of key characteristics, such as the entropy, log-probability, and their gradients, becomes intractable under complex constraints. Hence, prior work approximates these using the non-truncated distributions, which severely degrades performance. We argue that accurate estimation of these characteristics is crucial in the action-constrained RL setting, and propose efficient numerical approximations for them. We also provide an efficient sampling strategy for truncated policy distributions and validate our approach on three benchmark environments, which demonstrate significant performance improvements when using accurate estimations.

Zitieren

APA

Stolz, R., Eichelbeck, M., & Althoff, M. (2026). Improving Stochastic Action-Constrained Reinforcement Learning via Truncated Distributions. Proceedings of the AAAI Conference on Artificial Intelligence, 40(30), 25617-25626. https://doi.org/10.1609/aaai.v40i30.39758

Vancouver

Stolz R, Eichelbeck M, Althoff M. Improving Stochastic Action-Constrained Reinforcement Learning via Truncated Distributions. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(30):25617-25626. doi:10.1609/aaai.v40i30.39758

BibTeX
@article{Stolz2026,
  author = {Stolz, Roland and Eichelbeck, Michael and Althoff, Matthias},
  title = {Improving Stochastic Action-Constrained Reinforcement Learning via Truncated Distributions},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {30},
  pages = {25617--25626},
  doi = {10.1609/aaai.v40i30.39758}
}
Kostenlos starten →