ResearchBenchmarksMulti-Agent Navigation on 3-Agent 6-Landmark Medium Expert Standard EnvironmentFollow19.2Number of Good PoliciesDOM2-0.1444.8789.914.922Jul 4, 2023Evaluation ResultsMethodMethodLinksNumber of Good PoliciesDOM22023.0719.2MA-SfBC2023.0711.6OMAR2023.075.2MA-CQL2023.070.6