Reinforcement Learning on Hopper delta=[0.2, 0.5, 0.5], kappa=2.5 v5 (test)
3,312ReturnDD-SRad
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DD-SRadBackbone=TD32026.05 | 3,312 | 11.6 | 46.2 | |
| BoxPre+Backbone=TD32026.05 | 3,256 | 6 | 30.3 | |
| D-TanhBackbone=TD32026.05 | 2,712 | 8.4 | 49.6 | |
| DD-SRadBackbone=SAC2026.05 | 2,610 | 7 | 46.1 | |
| SRad-StrictBackbone=TD32026.05 | 2,394 | 13.4 | 16.6 | |
| D-TanhBackbone=SAC2026.05 | 2,334 | 11.8 | 48 | |
| BoxPre+Backbone=SAC2026.05 | 2,319 | 9 | 26.7 | |
| SRad-StrictBackbone=SAC2026.05 | 2,195 | 33.5 | 15.5 | |
| SRad-QPBackbone=SAC2026.05 | 1,624 | 20.4 | 37 | |
| Post(QP)Backbone=SAC2026.05 | 1,278 | 12.9 | 35.3 | |
| SRad-QPBackbone=TD32026.05 | 1,020 | 13.6 | 33.3 | |
| Post(QP)Backbone=TD32026.05 | 849 | 29.6 | 48.7 |