Unsupervised Reinforcement Learning on URL Benchmark (Walker)
237Flip ScoreRND
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| RNDmode=Zero-shot2022.10 | 237 | 89 | 392 | 195 | |
| APTmode=Zero-shot2022.10 | 235 | 91 | 401 | 204 | |
| CICmode=Zero-shot2022.10 | 218 | 79 | 356 | 174 | |
| Disagreementmode=Zero-shot2022.10 | 207 | 78 | 366 | 167 | |
| Protomode=Zero-shot2022.10 | 171 | 72 | 322 | 162 | |
| MOSSmode=Zero-shot2022.10 | 166 | 58 | 295 | 112 | |
| SMMmode=Zero-shot2022.10 | 117 | 49 | 229 | 100 | |
| ICMmode=Zero-shot2022.10 | 78 | 32 | 170 | 60 | |
| AS-BOBmode=Zero-shot2022.10 | 35 | 22 | 132 | 28 | |
| APSmode=Zero-shot2022.10 | 35 | 27 | 162 | 30 | |
| AS-ALICEmode=Zero-shot2022.10 | 31 | 24 | 139 | 28 | |
| DIAYNmode=Zero-shot2022.10 | 22 | 18 | 107 | 20 |