README.md

Safe Value Functions: Learned Critics as Hard Safety Constraints

This paper was awarded Outstanding Paper at WFVML 2023

This is the code for "Safe Value Functions: Learned Critics as Hard Safety Constraints"

• Project Page
• Paper
• Video

This codebase is implemented using PyTorch, building on CleanRL.

Experiments

Instructions to reproduce the experiments in the paper can be found here

Citing

If you find our work useful, please consider citing:

@article{
    tan2023value,
    title={Safe Value Functions: Learned Critics as Hard Safety Constraints}, 
    author={Daniel C. H. Tan and Fernando Acero and Robert McCarthy and Andromachi Maria Delfaki and Zhibin Li and Dimitrios Kanoulas},
    year={2024},
    eprint={2306.04026},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}