Language-guided multi-agent reinforcement learning

This project explored using language models to guide how agents explore during multi-agent reinforcement learning. The goal was to improve sample efficiency while agents learn to coordinate.

Explore the code · Watch the video.