Skip to yearly menu bar Skip to main content


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

Keywords: [ Relational Learning ] [ Algorithms ] [ Representation Learning ] [ Reinforcement Learning and Planning ] [ Multi-Agent RL ]


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.

Chat is not available.