Offline Reinforcement Learning on Atari Qbert 1% DQN-replay v0
104.2Gamer-normalized ScoreCQL
Evaluation Results
| Method | Links | |
|---|---|---|
| CQLdataset variant=1% DQN-replay2021.06 | 104.2 | |
| Behavior Cloningdataset variant=1% DQN-replay2021.06 | 17.3 | |
| Decision Transformerdataset variant=1% DQN-replay2021.06 | 15.4 | |
| QR-DQNdataset variant=1% DQN-replay2021.06 | 0 | |
| REMdataset variant=1% DQN-replay2021.06 | 0 |