Preprint · 2018 · Open access

A Lyapunov-based Approach to Safe Reinforcement Learning

Yinlam Chow, Ofir Nachum, Edgar A. Duéñez‐Guzmán, Mohammad Ghavamzadeh

arXiv (Cornell University), 31, 8092–8101 · Published 20 May 2018 · DOI 10.48550/arxiv.1805.07708

Open PDFView at the publisherCite this paper

In brief

“A Lyapunov-based Approach to Safe Reinforcement Learning” is a 2018 preprint by Yinlam Chow and 3 co-authors, published in arXiv (Cornell University). OpenAlex counts 688 citations as of 22 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).

Citations
688
Most cited in
2026 (171)
References
36
Authors
4

Abstract

In many real-world reinforcement learning (RL) problems, besides optimizing the main objective function, an agent must concurrently avoid violating a number of constraints. In particular, besides optimizing performance it is crucial to guarantee the safety of an agent during training as well as deployment (e.g. a robot should avoid taking actions - exploratory or not - which irrevocably harm its hardware). To incorporate safety in RL, we derive algorithms under the framework of constrained Markov decision problems (CMDPs), an extension of the standard Markov decision problems (MDPs) augmented with constraints on expected cumulative costs. Our approach hinges on a novel \emph{Lyapunov} method. We define and present a method for constructing Lyapunov functions, which provide an effective way to guarantee the global safety of a behavior policy during training via a set of local, linear constraints. Leveraging these theoretical underpinnings, we show how to use the Lyapunov approach to systematically transform dynamic programming (DP) and RL algorithms into their safe counterparts. To illustrate their effectiveness, we evaluate these algorithms in several CMDP planning and decision-making tasks on a safety benchmark domain. Our results show that our proposed method significantly outperforms existing baselines in balancing constraint satisfaction and performance.

Cite this paper

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

APA 7

Chow, Y., Nachum, O., Duéñez‐Guzmán, E. A., & Ghavamzadeh, M. (2018). A Lyapunov-based Approach to Safe Reinforcement Learning. arXiv (Cornell University), 31, 8092–8101. https://doi.org/10.48550/arxiv.1805.07708

MLA 9

Chow, Yinlam, et al. “A Lyapunov-based Approach to Safe Reinforcement Learning.” arXiv (Cornell University), vol. 31, 2018, pp. 8092–8101. https://doi.org/10.48550/arxiv.1805.07708.

Chicago

Chow, Yinlam, Ofir Nachum, Edgar A. Duéñez‐Guzmán, and Mohammad Ghavamzadeh. 2018. “A Lyapunov-based Approach to Safe Reinforcement Learning.” arXiv (Cornell University) 31: 8092–8101. https://doi.org/10.48550/arxiv.1805.07708.

Harvard

Chow, Y. et al. (2018) ‘A Lyapunov-based Approach to Safe Reinforcement Learning’, arXiv (Cornell University), 31, pp. 8092–8101. doi:10.48550/arxiv.1805.07708.

IEEE

Y. Chow, O. Nachum, E. A. Duéñez‐Guzmán, and M. Ghavamzadeh, “A Lyapunov-based Approach to Safe Reinforcement Learning,” arXiv (Cornell University), vol. 31, pp. 8092–8101, 2018, doi: 10.48550/arxiv.1805.07708.

BibTeX

@misc{chow2018lyapunov,
  title     = {{A Lyapunov-based Approach to Safe Reinforcement Learning}},
  author    = {Chow, Yinlam and Nachum, Ofir and Duéñez‐Guzmán, Edgar A. and Ghavamzadeh, Mohammad},
  howpublished = {arXiv (Cornell University)},
  year      = {2018},
  volume    = {31},
  pages     = {8092--8101},
  doi       = {10.48550/arxiv.1805.07708},
  url       = {https://doi.org/10.48550/arxiv.1805.07708}
}

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

Citations by year

YearCitationsShare of the best year
201844
20192929
20205959
20218080
20226767
20239393
20249090
20259595
2026171171

Topic and keywords

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

  • safe reinforcement learning
  • Lyapunov function
  • constrained Markov decision process
  • safety constraints
  • constraint satisfaction
  • dynamic programming
  • safety benchmark

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?

Chow, Y., Nachum, O., Duéñez‐Guzmán, E. A., & Ghavamzadeh, M. (2018). A Lyapunov-based Approach to Safe Reinforcement Learning. arXiv (Cornell University), 31, 8092–8101. https://doi.org/10.48550/arxiv.1805.07708

How many times has this paper been cited?

688 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?

Yinlam Chow, Ofir Nachum, Edgar A. Duéñez‐Guzmán, Mohammad Ghavamzadeh.

What is the DOI of this paper?

10.48550/arxiv.1805.07708, permanently at https://doi.org/10.48550/arxiv.1805.07708. 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.