MDPax: GPU-accelerated MDP solvers in Python with JAX

Python Submitted 01 August 2025Published 31 October 2025
Review

Editor: @diehlpk (all papers)
Reviewers: @arrrunn (all reviews), @thchang (all reviews)

Authors

Joseph Farrington (0000-0003-4156-3419), Wai Keong Wong (0000-0002-5742-0108), Kezhi Li (0000-0003-3073-3128), Martin Utley (0000-0001-9928-1516)

Citation

Farrington et al., (2025). MDPax: GPU-accelerated MDP solvers in Python with JAX. Journal of Open Source Software, 10(114), 8897, https://doi.org/10.21105/joss.08897

@article{Farrington2025, doi = {10.21105/joss.08897}, url = {https://doi.org/10.21105/joss.08897}, year = {2025}, publisher = {The Open Journal}, volume = {10}, number = {114}, pages = {8897}, author = {Farrington, Joseph and Wong, Wai Keong and Li, Kezhi and Utley, Martin}, title = {MDPax: GPU-accelerated MDP solvers in Python with JAX}, journal = {Journal of Open Source Software} }
Copy citation string · Copy BibTeX  
Tags

optimization dynamic programming reinforcement learning

Altmetrics
Markdown badge

 

License

Authors of JOSS papers retain copyright.

This work is licensed under a Creative Commons Attribution 4.0 International License.

Creative Commons License

Table of Contents
Public user content licensed CC BY 4.0 unless otherwise specified.
ISSN 2475-9066