ResearchBenchmarksMeta-Reinforcement Learning on Ant-Dir (OOD)Meta-Reinforcement Learning on Ant-Dir (OOD)Follow59Average ReturnMetaSTAR-2.3613.5729.545.43May 30, 2026Evaluation resultsMethodMethodLinksAverage ReturnMetaSTAREpisode=FinalEpisode=Final2026.0559CORROEpisode=FinalEpisode=Final2026.0540CSROEpisode=FinalEpisode=Final2026.057UNICORNEpisode=FinalEpisode=Final2026.051FOCALEpisode=FinalEpisode=Final2026.050