Unsupervised Reinforcement Learning on URL Benchmark Quadruped
627Jump ScoreMOSS
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| MOSSmode=Zero-shot2022.10 | 627 | 370 | 767 | 306 | |
| RNDmode=Zero-shot2022.10 | 549 | 319 | 638 | 321 | |
| Disagreementmode=Zero-shot2022.10 | 464 | 269 | 534 | 273 | |
| CICmode=Zero-shot2022.10 | 415 | 248 | 493 | 250 | |
| APTmode=Zero-shot2022.10 | 304 | 194 | 384 | 198 | |
| ICMmode=Zero-shot2022.10 | 225 | 150 | 299 | 153 | |
| AS-ALICEmode=Zero-shot2022.10 | 195 | 131 | 264 | 127 | |
| AS-BOBmode=Zero-shot2022.10 | 170 | 111 | 223 | 107 | |
| DIAYNmode=Zero-shot2022.10 | 136 | 91 | 181 | 93 | |
| APSmode=Zero-shot2022.10 | 134 | 89 | 180 | 89 | |
| SMMmode=Zero-shot2022.10 | 73 | 46 | 91 | 48 | |
| Protomode=Zero-shot2022.10 | 36 | 23 | 46 | 23 |