We propose Multi Agent Reflective Policy Optimization (MARPO) to alleviate the issue of sample inefficiency in multi agent reinforcement learning. MARPO consists of two key components: a reflection mechanism that leverages subsequent trajectories to enhance sample efficiency, and an asymmetric clipping mechanism that is derived from the KL divergence and dynamically adjusts the clipping range to improve training stability. We evaluate MARPO in classic multi agent environments, where it consistently outperforms other methods.
@article{arxiv.2512.22832,
title = {MARPO: A Reflective Policy Optimization for Multi Agent Reinforcement Learning},
author = {Cuiling Wu and Yaozhong Gan and Junliang Xing and Ying Fu},
journal= {arXiv preprint arXiv:2512.22832},
year = {2025}
}