Pick and Place on Real-world Tabletop Manipulation
81.3Success Rate (SR)RL-Co
Evaluation Results
| Method | Links | |
|---|---|---|
| RL-CoVLA Model=π0.52026.02 | 81.3 | |
| Real-Only TrainingVLA Model=π0.52026.02 | 71.9 | |
| SFT Co-TrainingVLA Model=π0.52026.02 | 68.8 | |
| RL-CoVLA Model=OpenVLA2026.02 | 58.8 | |
| SFT Co-TrainingVLA Model=OpenVLA2026.02 | 23.4 | |
| Real-Only TrainingVLA Model=OpenVLA2026.02 | 6.3 |