Goal-Conditioned Reinforcement Learning on OGBench humanoidmaze-giant-navigate v0
92.8Success RateOTA +TTGS
Evaluation Results
| Method | Links | |
|---|---|---|
| OTA +TTGSMode=TTGS Augmented2025.10 | 92.8 | |
| OTAMode=Base Learner2025.10 | 89.1 | |
| GASMode=Planning Method2025.10 | 85.7 | |
| HIQL +TTGSMode=TTGS Augmented2025.10 | 85.3 | |
| SAW +TTGSMode=TTGS Augmented2025.10 | 79 | |
| SAWMode=Base Learner2025.10 | 40.4 | |
| HIQLMode=Base Learner2025.10 | 16 | |
| QRLMode=Base Learner2025.10 | 1.1 | |
| QRL +TTGSMode=TTGS Augmented2025.10 | 1 | |
| GCIQLMode=Base Learner2025.10 | 0.7 | |
| GCIQL +TTGSMode=TTGS Augmented2025.10 | 0.5 |