Offline Reinforcement Learning on AntMaze (large-play (l-p))
70Normalized ScoreQCS-G
Evaluation Results
| Method | Links | |
|---|---|---|
| QCS-GAlgorithm Category=Ours, Conditioning Strategy=Goal2024.02 | 70 | |
| QCS-RAlgorithm Category=Ours, Conditioning Strategy=Return2024.02 | 68.7 | |
| PORAlgorithm Category=Combined2024.02 | 58 | |
| SQLAlgorithm Category=Value-Based Method2024.02 | 53.2 | |
| IQLAlgorithm Category=Value-Based Method2024.02 | 39.6 | |
| RvS-GAlgorithm Category=RCSL, Conditioning Strategy=Goal2024.02 | 32.4 | |
| CQLAlgorithm Category=Value-Based Method2024.02 | 15.8 | |
| DCAlgorithm Category=RCSL2024.02 | 4.8 | |
| RvS-RAlgorithm Category=RCSL, Conditioning Strategy=Return2024.02 | 3.5 | |
| TD3+BCAlgorithm Category=Value-Based Method2024.02 | 0.2 | |
| DTAlgorithm Category=RCSL2024.02 | 0 |