Preprint · 2025 · Open access
SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints
arXiv (Cornell University) · Published 29 December 2025 · DOI 10.48550/arxiv.2512.23770
In brief
“SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints” is a 2025 preprint by Dominik Wagner and 2 co-authors, published in arXiv (Cornell University). OpenAlex counts 1 citations as of 22 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).
- Citations
- 1
- Most cited in
- 2026 (1)
- References
- 28
- Authors
- 3
Abstract
In safety-critical domains, reinforcement learning (RL) systems must satisfy strict, zero-cost safety constraints while achieving meaningful task performance. Existing model-free methods can struggle to achieve high safety without substantially compromising task performance. We introduce \emph{Safety-Biased Trust Region Policy Optimisation (SB-TRPO)}, a principled approach to RL with zero-cost constraints, which requires only a fixed fraction of the maximal cost reduction achievable within the trust region, thus retaining flexibility for reward optimisation. We show that the idealised update nevertheless converges to zero cost and maximal reward amongst zero-cost policies in finite MDPs. A practical gradient-based approximation provides local improvements in both safety and reward under suitable gradient alignment. Experiments on \emph{Safety Gymnasium} demonstrate high safety alongside strong task performance.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Wagner, D., Kanwar, A., & Ong, C.-H. L. (2025). SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2512.23770
MLA 9
Wagner, Dominik, et al. “SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints.” arXiv (Cornell University), 2025. https://doi.org/10.48550/arxiv.2512.23770.
Chicago
Wagner, Dominik, Ankit Kanwar, and C.-H. Luke Ong. 2025. “SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2512.23770.
Harvard
Wagner, D., Kanwar, A. and Ong, C.-H. L. (2025) ‘SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints’, arXiv (Cornell University). doi:10.48550/arxiv.2512.23770.
IEEE
D. Wagner, A. Kanwar, and C.-H. L. Ong, “SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints,” arXiv (Cornell University), 2025, doi: 10.48550/arxiv.2512.23770.
BibTeX
@misc{wagner2025trpo,
title = {{SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints}},
author = {Wagner, Dominik and Kanwar, Ankit and Ong, C.-H. Luke},
howpublished = {arXiv (Cornell University)},
year = {2025},
doi = {10.48550/arxiv.2512.23770},
url = {https://doi.org/10.48550/arxiv.2512.23770}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Topic and keywords
Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- safe reinforcement learning
- trust region policy optimization
- safety constraints
- safety-critical systems
Work with this paper
Most-cited papers in Computer Science
- ImageNet classification with deep convolutional neural networksAlex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · Communications of the ACM · 2017 · 114,726 citations
- Long Short-Term MemorySepp Hochreiter, Jürgen Schmidhuber · Neural Computation · 1997 · 109,097 citations
- A Mathematical Theory of CommunicationClaude E. Shannon · Bell System Technical Journal · 1948 · 91,736 citations
- Deep learningYann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature · 2015 · 90,226 citations
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksShaoqing Ren, Kaiming He, Ross Girshick and 1 more · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2016 · 74,143 citations
Questions about this paper
How do I cite this paper in APA style?
Wagner, D., Kanwar, A., & Ong, C.-H. L. (2025). SB-TRPO: Towards Safe Reinforcement Learning with Hard Constraints. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2512.23770
How many times has this paper been cited?
1 times, according to OpenAlex on 22 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.
Who wrote this paper?
Dominik Wagner, Ankit Kanwar, C.-H. Luke Ong.
What is the DOI of this paper?
10.48550/arxiv.2512.23770, permanently at https://doi.org/10.48550/arxiv.2512.23770. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 22 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.