# CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guarantee

> Research article (arXiv (Cornell University), 2020) · cited 16× · AI/ML

**Wikidata**: [openalex:W3168269639](https://www.wikidata.org/wiki/openalex:W3168269639)  
**Source**: https://4ort.xyz/entity/crpo-a-new-approach-for-safe-reinforcement-learning-with-convergence-guarantee