Safe Multi-agent Learning via Trapping Regions

Open in new window