Goal-Conditioned Reinforcement Learning on OGBench pointmaze-giant-stitch v0
98Success RateGCIQL +TTGS
Evaluation Results
| Method | Links | |
|---|---|---|
| GCIQL +TTGSMode=TTGS Augmented2025.10 | 98 | |
| QRL +TTGSMode=TTGS Augmented2025.10 | 93.2 | |
| OTA +TTGSMode=TTGS Augmented2025.10 | 86.8 | |
| SAW +TTGSMode=TTGS Augmented2025.10 | 82.3 | |
| HIQL +TTGSMode=TTGS Augmented2025.10 | 80.9 | |
| CDMode=Planning Method2025.10 | 68 | |
| QRLMode=Base Learner2025.10 | 55.3 | |
| OTAMode=Base Learner2025.10 | 42.9 | |
| SAWMode=Base Learner2025.10 | 6.8 | |
| HIQLMode=Base Learner2025.10 | 0 | |
| GCIQLMode=Base Learner2025.10 | 0 |