Goal-conditioned Reinforcement Learning on OGBench antmaze large navigate (5 tasks) zero-shot
34Avg ReturnHILP
Evaluation Results
| Method | Links | |
|---|---|---|
| HILPevaluation_mode=Zero-shot evaluation2026.02 | 34 | |
| One-Step FBevaluation_mode=Zero-shot evaluation2026.02 | 30 | |
| FBevaluation_mode=Zero-shot evaluation2026.02 | 25 | |
| ICVFevaluation_mode=Zero-shot evaluation2026.02 | 23 | |
| BYOL-gammaevaluation_mode=Zero-shot evaluation2026.02 | 21 | |
| Laplacianevaluation_mode=Zero-shot evaluation2026.02 | 9 |