Visual Offline Reinforcement Learning on V-D4RL (various)
524Cheetah-Run MediumDRIML
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| DRIML2023.10 | 524 | 425 | 395 | 235 | 403 | 399 | 310 | 427 | 3,118 | |
| SimSR+RS+EBSreward scaling=true, expectile-based operator=true2023.10 | 491 | 480 | 462 | 240 | 547 | 845 | 454 | 580 | 4,043 | |
| HOMER2023.10 | 475 | 439 | 306 | 283 | 383 | 781 | 218 | 686 | 3,571 | |
| MICo+EBSexpectile-based operator=true2023.10 | 449 | 447 | 357 | 240 | 341 | 635 | 331 | 447 | 3,253 | |
| CURL2023.10 | 392 | 452 | 271 | 265 | 348 | 729 | 200 | 769 | 3,426 | |
| SimSR2023.10 | 391 | 443 | 374 | 197 | 360 | 755 | 409 | 578 | 3,507 | |
| ICM2023.10 | 365 | 358 | 251 | 167 | 280 | 606 | 237 | 850 | 3,114 | |
| MICo2023.10 | 177 | 450 | 335 | 207 | 282 | 586 | 308 | 370 | 2,715 |