ResearchBenchmarksOffline Reinforcement Learning on Maze2D umaze v0 (test)Follow111Overall ScoreOptiDICE-0.90428.14857.286.252Oct 10, 2023Oct 30, 2023Nov 19, 2023Dec 9, 2023Dec 29, 2023Jan 18, 2024Feb 7, 2024Evaluation ResultsMethodMethodLinksOverall ScoreOptiDICESeeds=5Seeds=52023.10111ProposedSeeds=5Seeds=52023.1096.5ATACSeeds=5Seeds=52023.1084.4CQLSeeds=5Seeds=52023.1050.5IQLSeeds=5Seeds=52023.1041.5COMBOSeeds=5Seeds=52023.1034.2TD3+BCSeeds=5Seeds=52023.1013.8BCQSeeds=5Seeds=52023.1012.8CQL2024.025.7BEARSeeds=5Seeds=52023.103.4