Preprint · 2021 · Open access

Safe Reinforcement Learning Using Advantage-Based Intervention

Nolan Wagener, Boots, Byron, Cheng, Ching-An

arXiv (Cornell University) · Published 16 June 2021 · DOI 10.48550/arxiv.2106.09110

Open PDFView at the publisherCite this paper

In brief

“Safe Reinforcement Learning Using Advantage-Based Intervention” is a 2021 preprint by Nolan Wagener and 2 co-authors, published in arXiv (Cornell University). OpenAlex counts 50 citations as of 22 November 2026. It is open access under a CC BY licence: the full text is free to read at arXiv (Cornell University).

Citations
50
Most cited in
2023 (17)
References
43
Authors
3

Abstract

Many sequential decision problems involve finding a policy that maximizes total reward while obeying safety constraints. Although much recent research has focused on the development of safe reinforcement learning (RL) algorithms that produce a safe policy after training, ensuring safety during training as well remains an open problem. A fundamental challenge is performing exploration while still satisfying constraints in an unknown Markov decision process (MDP). In this work, we address this problem for the chance-constrained setting. We propose a new algorithm, SAILR, that uses an intervention mechanism based on advantage functions to keep the agent safe throughout training and optimizes the agent's policy using off-the-shelf RL algorithms designed for unconstrained MDPs. Our method comes with strong guarantees on safety during both training and deployment (i.e., after training and without the intervention mechanism) and policy performance compared to the optimal safety-constrained policy. In our experiments, we show that SAILR violates constraints far less during training than standard safe RL and constrained MDP approaches and converges to a well-performing policy that can be deployed safely without intervention. Our code is available at https://github.com/nolanwagener/safe_rl.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Wagener, N., Byron, B., & Ching-An, C. (2021). Safe Reinforcement Learning Using Advantage-Based Intervention. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2106.09110

MLA 9

Wagener, Nolan, et al. “Safe Reinforcement Learning Using Advantage-Based Intervention.” arXiv (Cornell University), 2021. https://doi.org/10.48550/arxiv.2106.09110.

Chicago

Wagener, Nolan, Boots, Byron, and Cheng, Ching-An. 2021. “Safe Reinforcement Learning Using Advantage-Based Intervention.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2106.09110.

Harvard

Wagener, N., Byron, B. and Ching-An, C. (2021) ‘Safe Reinforcement Learning Using Advantage-Based Intervention’, arXiv (Cornell University). doi:10.48550/arxiv.2106.09110.

IEEE

N. Wagener, B. Byron, and C. Ching-An, “Safe Reinforcement Learning Using Advantage-Based Intervention,” arXiv (Cornell University), 2021, doi: 10.48550/arxiv.2106.09110.

BibTeX

@misc{wagener2021safe,
  title     = {{Safe Reinforcement Learning Using Advantage-Based Intervention}},
  author    = {Wagener, Nolan and Byron, Boots, and Ching-An, Cheng,},
  howpublished = {arXiv (Cornell University)},
  year      = {2021},
  doi       = {10.48550/arxiv.2106.09110},
  url       = {https://doi.org/10.48550/arxiv.2106.09110}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
202133
202244
20231717
20241212
20251212
202622

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics

  • safe reinforcement learning
  • constrained Markov decision process
  • safety guarantees
  • policy optimization

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Wagener, N., Byron, B., & Ching-An, C. (2021). Safe Reinforcement Learning Using Advantage-Based Intervention. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2106.09110

How many times has this paper been cited?

50 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?

Nolan Wagener, Boots, Byron, Cheng, Ching-An.

What is the DOI of this paper?

10.48550/arxiv.2106.09110, permanently at https://doi.org/10.48550/arxiv.2106.09110. 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.