Offline Reinforcement Learning on Antmaze umaze-diverse benchmark leaderboard · SOTA2 Research

Offline Reinforcement Learning on Antmaze umaze-diverse

90.7Average ReturnRORL

Evaluation Results

Method	Links
RORL2022.06		90.7
CQL2024.02		84
CQL2022.06		84
CQL-Min2024.01		84
SPQR2024.01		80
TD3+BC2022.06		71.4
IQL+smoothing2022.06		64
IQL2022.06		62.2
IQL2024.01		62.2
ROMI+BCQ2022.06		61.2
BEAR2024.01		61
BC2024.01		55
AWAC2022.06		49.3
BC2022.06		45.6
SAC-Min2024.01		0