List your company

Offline Reinforcement Learning on D4RL walker2d-random

510Normalized ScoreAWAC
-20.4117.3255392.7Jun 12, 2021Apr 6, 2022Jan 30, 2023Nov 24, 2023Sep 18, 2024Jul 13, 2025May 8, 2026

Evaluation results

MethodLinks
AWACimplementation=reported2021.06510
CQLimplementation=author-provided (re-run)2021.06270
TD3+BCalpha=2.52021.06140
BCimplementation=own2021.06120
BRAC-pimplementation=reported2021.0680
Fisher-BRCimplementation=author-provided (re-run)2021.0660
MOREL2021.1037.3
BC2021.1036
NEUBAYMethod Category=Ours2025.1234.1
SUMOMethod Category=Conservative model-based2025.1227.9
MoMoMethod Category=Conservative model-based2025.1226.8
RRPINumber of random seeds=42026.0323.7
MoDAPMethod Category=Bayesian-inspired2025.1223.1
EPQNumber of random seeds=42026.0323
EPQ2026.0523
ADMPOMethod Category=Conservative model-based2025.1222.2
AMGNumber of random seeds=42026.0322.2
ADM2026.0522.2
MAPLEMethod Category=Bayesian-inspired2025.1222.1
PSPOAveraged across=4 random seeds2026.0522.1
Count-MORLMethod Type=Model-based2023.1021.9
PMDBNumber of random seeds=42026.0321.8
PMDBMethod Type=Model-based2023.1021.8
PMDB2026.0521.8
SAC-NImplementation=Ours2021.1021.7
RAMBOMethod Type=Model-based2023.1021.6
LEQMethod Category=Conservative model-based2025.1221.5
RORLImplementation Source=Ours, Ensemble Size=102022.0621.4
CROPMethod Type=Model-based2023.1020.9
VIPOMethod Category=Conservative model-based2025.1220
MOBILEMethod Category=Conservative model-based2025.1217.9
COMBOMethod Type=Model-based2023.1017.9
CBOPMethod Category=Bayesian-inspired2025.1217.8
EDACImplementation=Ours2021.1016.6
EDACImplementation Source=Paper2022.0616.6
EDACMethod Category=Model-free2025.1216.6
EDACMethod Type=Model-free2023.1016.6
MOReLNumber of random seeds=42026.0316
MOReL2026.0516
APE-VMethod Category=Conservative model-based2025.1215.5
RAMBOMethod Category=Model-free2025.1211.5
ProposedSeeds=52023.1011.2
BEAR2021.1011.1
BEARalgorithm class=model-free2021.1011.1
OptiDICESeeds=52023.109.9
PBRL2022.068.1
ROMI-CQL2021.107.5
ROMI-CQLbase algorithm=CQL2021.107.5
MOPOMethod Category=Conservative model-based2025.127.4
BEARSeeds=52023.107.3
Fu et al.Algorithmic Template=Iterative2021.067.3
COMBOSeeds=52023.107
CQLImplementation=Original Paper2021.107
COMBO2021.107
COMBOalgorithm class=model-based2021.107
COMBOMethod Category=Model-free2025.127
REMImplementation=Reproduced2021.106.9
ATACSeeds=52023.106.8
ATACMethod Type=Model-free2023.106.8
EDAC-10Implementation Source=Reproduced, Ensemble Size=102022.066.7
CABI+TD3-BCMethod Type=Model-based2023.106.4
Rev. KL Reg.Algorithmic Template=One-step2021.066.1
IQLSeeds=52023.105.8
TT2022.065.6
CQLImplementation=Reproduced2021.105.4
CQL2022.065.4
CQLMethod Category=Model-free2025.125.4
CQLNumber of random seeds=42026.035.4
CQL2026.055.4
Exp. WeightAlgorithmic Template=One-step2021.065.2
BooT-oGeneration Strategy=Autoregressive2022.065.2
ROMI-BCQ2021.105
BCQSeeds=52023.104.9
BCQImplementation=Reproduced2021.104.9
BooT-rGeneration Strategy=Autoregressive2022.064.8
BooT-rGeneration Strategy=Teacher-forcing2022.064.8
DMGNumber of random seeds=42026.034.8
DMG2026.054.8
CQLSeeds=52023.104.7
BooT-oGeneration Strategy=Teacher-forcing2022.064.6
BEARImplementation=Reproduced2021.104.3
MOPO2021.104.2
Easy BCQAlgorithmic Template=One-step2021.064.2
TD3+BCSeeds=52023.103.4
CQL2021.102.4
BAIL2021.102.4
SAC-10Implementation Source=Reproduced, Ensemble Size=102022.061.5
BCImplementation=Reproduced2021.101.3
UWACImplementation=Reproduced2021.101.3
BC2022.061.3
BCQ2021.101.1
SACImplementation=Reproduced2021.100.9
BCAlgorithmic Template=One-step2021.060.7
MORELImplementation=Reproduced2021.100.2
BRAC-P2021.100
BRAC-v2021.100
Repb-SDE2021.100
BC2021.100
RAMBONumber of random seeds=42026.030
RAMBO2026.050