Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

MinAtar

Benchmarks

Task NameDataset NameSOTA ResultTrend
Reinforcement LearningMinAtar
Freeway Score71.34
11
Offline-to-online Reinforcement LearningMinAtar
Breakout Score (Online)17.1
10
Reinforcement LearningMinAtar 5M frames
Mean Return52.7
9
Reinforcement LearningMinAtar Breakout (test)
Mean Episodic Reward25.77
6
Reinforcement LearningMinAtar Asterix (test)
Mean Episodic Reward19.45
6
Reinforcement LearningMinAtar SpaceInvaders (test)
Mean Episodic Reward101.42
6
Reinforcement LearningMinAtar Seaquest (test)
Mean Episodic Reward27.58
6
Continual Reinforcement LearningMinAtar
Breakout Score14.54
6
Online Reinforcement LearningMinAtar (|A|=216, k=3) Macro-Action (online)
Breakout Score11.98
5
Reinforcement LearningMinAtar (test)
Asterix Score16.16
4
Reinforcement LearningMinAtar
Overall Score5
4
Reinforcement LearningMinAtar SpaceInvaders
Metric-
0
Reinforcement LearningMinAtar Seaquest
Metric-
0
Reinforcement LearningMinAtar Freeway
Metric-
0
Reinforcement LearningMinAtar Breakout
Metric-
0
Reinforcement LearningMinAtar Asterix
Metric-
0
Showing 16 of 16 rows