Continuous Control on Humanoid v5
5,906.7Average ReturnDBC
Evaluation Results
| Method | Links | |
|---|---|---|
| DBCActor Backbone=SAC2026.02 | 5,906.7 | |
| TRFP(ours)NFE=4 × 42026.04 | 5,370.8 | |
| SACNFE=12026.04 | 5,350.5 | |
| TRFP(one-step)NFE=1 × 42026.04 | 5,332.5 | |
| TQCActor Backbone=SAC2026.02 | 5,269 | |
| SACActor Backbone=SAC2026.02 | 5,207.1 | |
| TD3NFE=12026.04 | 5,138.6 | |
| VFActor Backbone=SAC2026.02 | 4,950.3 | |
| VDActor Backbone=SAC2026.02 | 4,886.9 | |
| IQNActor Backbone=SAC2026.02 | 4,729.2 | |
| SDACNFE=20 × 322026.04 | 4,087.9 | |
| DSACActor Backbone=SAC2026.02 | 3,442.6 | |
| MaxEntDPNFE=20 × 102026.04 | 3,084.2 |