Sparse-reward manipulation on Coffee simulated environment
96Success RateCSIL++ Ens.
Evaluation Results
| Method | Links | |
|---|---|---|
| CSIL++ Ens.Policy architecture=Ensemble, Learning step=200k2026.06 | 96 | |
| CSIL++ Res.Policy architecture=Residual, Learning step=0k2026.06 | 96 | |
| XQC+OD Ens.Policy architecture=Ensemble, Learning step=0k2026.06 | 96 | |
| XQC+OD Res.Policy architecture=Residual, Learning step=0k2026.06 | 88 | |
| VLA2026.06 | 78 | |
| PLD RL (Res.)Policy architecture=Residual, Learning step=250k2026.06 | 68 |