ResearchBenchmarksReinforcement Learning on OCAtari AsterixFollow62,591,150ScoreNUDGE-2,503,516.5214,396,060.36531,295,637.2548,195,214.135Jun 2, 2023Evaluation ResultsMethodMethodLinksScoreNUDGEtraining_mode=expert s...training_mode=expert supervision2023.0662,591,150Random2023.06235,134DQN2023.06124.5