Multi-task Reinforcement Learning on Atari-57 clipped
110.7Median Human Normalized ScorePopArt-IMPALA
Evaluation Results
| Method | Links | |
|---|---|---|
| PopArt-IMPALAEvaluation Start Type=Random2018.09 | 110.7 | |
| PopArt-IMPALAEvaluation Start Type=Human2018.09 | 101.5 | |
| IMPALAEvaluation Start Type=Random2018.09 | 59.7 | |
| IMPALAEvaluation Start Type=Human2018.09 | 28.5 |