List your company
ResearchFormerly WizWand

Reinforcement Learning on Atari 2600 Montezuma's Revenge (test)

14,973ScoreSND-V
-598.923,443.797,486.511,529.21Mar 3, 2017Mar 1, 2018Feb 28, 2019Feb 27, 2020Feb 24, 2021Feb 23, 2022Feb 22, 2023
Updated 8mo ago

Evaluation results

MethodLinks
SND-V2023.0214,973
RND2018.108,152
SND-VIC2023.027,838
SND-STD2023.027,212
RND2023.026,689
Avg. Human2018.104,753
DQN-CTSTraining frames=150M, Seeds=32017.033,705.5
SOTAcitation=[2]2018.103,700
DQN-PixelCNNTraining frames=150M, Seeds=32017.032,514.3
A2C+SILSelf-Imitation Learning=true2018.062,500
PPO2018.102,497
Dynamics2018.10400
Baseline2023.02400
A3C-CTSTraining frames=150M, Seeds=32017.03273.7
A3C-CTSExploration=CTS2018.06273
Reactor-PixelCNNExploration=PixelCNN2018.06100
TRPO-AE-SimHashExploration=SimHash2018.0675
SND-V2023.028.87
SND-VIC2023.028.45
SND-STD2023.027.76
RND2023.025.33
DQNTraining frames=150M, Seeds=32017.030
Prior. DuelTraining frames=150M, Seeds=32017.030
Baseline2023.020