Goal-conditioned Reinforcement Learning on hopper medium-expert v2 (G2_balanced)
1.678Goal-Conditioned ReturnFrozen
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Frozen2026.04 | 1.678 | 0 | |
| KL-Reg2026.04 | 1.649 | 3.336 | |
| PoE2026.04 | 1.644 | 2.826 | |
| Additive2026.04 | 1.638 | 37.213 | |
| Prior Only2026.04 | 1.621 | 146.257 |