PIMAEX: Multi-Agent Exploration through Peer Incentivization

While exploration in single-agent reinforcement learning has been studied extensively in recent years, considerably less work has focused on its counterpart in multi-agent reinforcement learning. To address this issue, this work proposes a peer-incentivized reward function inspired by previous research on intrinsic curiosity and influence-based rewards. The textitPIMAEX reward, short for Peer-Incentivized Multi-Agent Exploration, aims to improve exploration in the multi-agent setting by encouraging agents to exert influence over each other to increase the likelihood of encountering novel states. We evaluate the textitPIMAEX reward in conjunction with textitPIMAEX-Communication, a multi-agent training algorithm that employs a communication channel for agents to influence one another. The evaluation is conducted in the textitConsume/Explore environment, a partially observable environment with deceptive rewards, specifically designed to challenge the exploration vs. exploitation dilemma and the credit-assignment problem. The results empirically demonstrate that agents using the textitPIMAEX reward with textitPIMAEX-Communication outperform those that do not.