A Policy Gradient Algorithm for Learning to Learn in Multiagent Reinforcement Learning



  • Dong-Ki Kim
  • Miao Liu
  • Matthew Riemer
  • Chuangchuang Sun
  • Marwa Abdulhai
  • Golnaz Habibi
  • Sebastian Lopez-Cot
  • Gerald Tesauro
  • Jonathan How

Published on


A fundamental challenge in multiagent reinforcement learning is to learn beneficial behaviors in a shared environment with other simultaneously learning agents. In particular, each agent perceives the environment as effectively non-stationary due to the changing policies of other agents. Moreover, each agent is itself constantly learning, leading to natural non-stationarity in the distribution of experiences encountered. In this paper, we propose a novel meta-multiagent policy gradient theorem that directly accounts for the non-stationary policy dynamics inherent to multiagent learning settings. This is achieved by modeling our gradient updates to consider both an agent’s own non-stationary policy dynamics and the non-stationary policy dynamics of other agents in the environment. We show that our theoretically grounded approach provides a general solution to the multiagent learning problem, which inherently comprises all key aspects of previous state of the art approaches on this topic. We test our method on a diverse suite of multiagent benchmarks and demonstrate a more efficient ability to adapt to new agents as they learn than baseline methods across the full spectrum of mixed incentive, competitive, and cooperative domains.

This paper has been published at ICML 2021

Please cite our work using the BibTeX below.

      title={A Policy Gradient Algorithm for Learning to Learn in Multiagent Reinforcement Learning}, 
      author={Dong-Ki Kim and Miao Liu and Matthew Riemer and Chuangchuang Sun and Marwa Abdulhai and Golnaz Habibi and Sebastian Lopez-Cot and Gerald Tesauro and Jonathan P. How},
Close Modal