Transformer Guided Coevolution: Improved Team Selection in Multiagent Adversarial Team Games

TL;DR

This paper introduces BERTeam, a transformer-based algorithm for optimal team selection in multiagent adversarial games, demonstrating superior performance in Marine Capture-The-Flag compared to existing methods.

Contribution

The paper presents a novel transformer-guided coevolutionary approach for team selection, integrating deep neural networks with reinforcement learning in multiagent adversarial settings.

Findings

BERTeam learns effective, non-trivial team compositions.

BERTeam outperforms MCAA in Marine Capture-The-Flag.

The approach generalizes well against unseen opponents.

Abstract

We consider the problem of team selection within multiagent adversarial team games. We propose BERTeam, a novel algorithm that uses a transformer-based deep neural network with Masked Language Model training to select the best team of players from a trained population. We integrate this with coevolutionary deep reinforcement learning, which trains a diverse set of individual players to choose from. We test our algorithm in the multiagent adversarial game Marine Capture-The-Flag, and find that BERTeam learns non-trivial team compositions that perform well against unseen opponents. For this game, we find that BERTeam outperforms MCAA, an algorithm that similarly optimizes team selection.