Offline Reinforcement Learning on D4RL maze2d-large (sparse) benchmark leaderboard · SOTA2 Research

Offline Reinforcement Learning on D4RL maze2d-large (sparse)

83.1Normalized ScoreROMI-BCQ

Evaluation Results

Method	Links
ROMI-BCQ2021.10		83.1
BEAR2021.10		81
BCQ2021.10		30.8
CQL2021.10		16
COMBO2021.10		14.1
ROMI-CQL2021.10		10.2
BAIL2021.10		4.4
BRAC-P2021.10		0.2
MOPO2021.10		-0.8
BC2021.10		-1.7
BRAC-v2021.10		-2.2
Repb-SDE2021.10		-2.2
MOREL2021.10		-2.3