Reinforcement Learning on Atari 2600 Venture ALE (test)
1,400ScoreReactor-PixelCNN
Evaluation results
| Method | Links | |
|---|---|---|
| Reactor-PixelCNNExploration=PixelCNN2018.06 | 1,400 | |
| DQN-PixelCNNTraining frames=150M, Seeds=32017.03 | 1,356.25 | |
| TRPO-AE-SimHashExploration=SimHash2018.06 | 445 | |
| DQNTraining frames=150M, Seeds=32017.03 | 163 | |
| DQN-CTSTraining frames=150M, Seeds=32017.03 | 82.2 | |
| Prior. DuelTraining frames=150M, Seeds=32017.03 | 48 | |
| A3C-CTSTraining frames=150M, Seeds=32017.03 | 0 | |
| A2C+SILSelf-Imitation Learning=true2018.06 | 0 | |
| A3C-CTSExploration=CTS2018.06 | 0 |