We introduce Multi-CALF, an algorithm that intelligently combines reinforcement learning policies based on their relative value improvements. Our approach integrates a standard RL policy with a theoretically-backed alternative policy, inheriting formal stability guarantees while often achieving better performance than either policy individually. We prove that our combined policy converges to a specified goal set with known probability and provide precise bounds on maximum deviation and convergence time. Empirical validation on control tasks demonstrates enhanced performance while maintaining stability guarantees.
@article{arxiv.2505.12350,
title = {Multi-CALF: A Policy Combination Approach with Statistical Guarantees},
author = {Georgiy Malaniya and Anton Bolychev and Grigory Yaremenko and Anastasia Krasnaya and Pavel Osinenko},
journal= {arXiv preprint arXiv:2505.12350},
year = {2025}
}