Announcement_3

We introduce GMFS, a scalable framework for multi-agent reinforcement learning that maintains near-optimal performance in heterogeneous populations.