Continuous Control on DeepMind Control Suite (Environment Returns)
873Cheetah RunDRND
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| DRNDEnsemble size=2, Update-to-Data Ratio=20, Interaction budget=300 0002025.07 | 873 | 3 | 251 | 1 | 213 | 166 | 224 | 184 | 689 | |
| DRNDEnsemble size=2, Update-to-Data Ratio=20, Interaction budget=300,000, Number of seeds=102025.07 | 871 | 9 | 374 | 1 | 204 | 158 | 234 | 200 | 688 | |
| REDQEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300 0002025.07 | 866 | 54 | 109 | 131 | 559 | 454 | 847 | 932 | 649 | |
| REDQEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300,000, Number of seeds=102025.07 | 851 | 105 | 318 | 117 | 539 | 416 | 728 | 878 | 631 | |
| AEAEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300 0002025.07 | 827 | 86 | 935 | 144 | 665 | 484 | 856 | 947 | 741 | |
| AEAEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300,000, Number of seeds=102025.07 | 823 | 130 | 779 | 145 | 678 | 478 | 823 | 935 | 705 | |
| DSACEnsemble size=2, Update-to-Data Ratio=20, Interaction budget=300 0002025.07 | 753 | 54 | 39 | 17 | 284 | 2 | 536 | 555 | 658 | |
| DSACEnsemble size=2, Update-to-Data Ratio=20, Interaction budget=300,000, Number of seeds=102025.07 | 746 | 79 | 273 | 17 | 271 | 2 | 520 | 570 | 650 | |
| REDQEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300 000, Number of seeds=102025.07 | 670 | 63 | 257 | 57 | 270 | 196 | 547 | 653 | 496 | |
| SAC-NEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300 0002025.07 | 635 | 0 | 0 | 1 | 9 | 2 | 43 | 112 | 258 | |
| AEAEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300 000, Number of seeds=102025.07 | 627 | 74 | 567 | 80 | 315 | 239 | 620 | 744 | 570 | |
| DRNDEnsemble size=2, Update-to-Data Ratio=20, Interaction budget=300 000, Number of seeds=102025.07 | 623 | 5 | 215 | 1 | 80 | 53 | 182 | 147 | 516 | |
| SAC-NEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300,000, Number of seeds=102025.07 | 596 | 1 | 5 | 1 | 52 | 2 | 209 | 284 | 270 | |
| DSACEnsemble size=2, Update-to-Data Ratio=20, Interaction budget=300 000, Number of seeds=102025.07 | 544 | 33 | 181 | 2 | 89 | 3 | 341 | 374 | 502 | |
| SAC-NEnsemble size=10, Update-to-Data Ratio=20, Interaction budget=300 000, Number of seeds=102025.07 | 466 | 0 | 4 | 1 | 27 | 2 | 191 | 231 | 188 |