Sparse-reward manipulation on Nut Assembly simulated environment
40Success RateCSIL++ Res.
Evaluation Results
| Method | Links | |
|---|---|---|
| CSIL++ Res.Policy architecture=Residual, Learning step=0k2026.06 | 40 | |
| XQC+OD Res.Policy architecture=Residual, Learning step=0k2026.06 | 34 | |
| XQC+OD Ens.Policy architecture=Ensemble, Learning step=0k2026.06 | 24 | |
| VLA2026.06 | 22 | |
| CSIL++ Ens.Policy architecture=Ensemble, Learning step=150k2026.06 | 22 | |
| PLD RL (Res.)Policy architecture=Residual, Learning step=0k2026.06 | 0 |