Failure Detection on Blocks Real-world
95AUCPRACC
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ACCPolicy=π0.52026.06 | 95 | 24 | |
| ACCPolicy=X-VLA2026.06 | 92 | 24 | |
| FAILPolicy=X-VLA2026.06 | 86 | 25 | |
| DiffPolicy=π0.52026.06 | 85 | 31 | |
| DiffPolicy=X-VLA2026.06 | 84 | 28 | |
| STACPolicy=π0.52026.06 | 84 | 36 | |
| LLMDPolicy=X-VLA2026.06 | 83 | 26 | |
| ACEPolicy=X-VLA2026.06 | 82 | 32 | |
| FAILPolicy=π0.52026.06 | 81 | 27 | |
| LLMDPolicy=π0.52026.06 | 73 | 31 | |
| ACEPolicy=π0.52026.06 | 60 | 45 | |
| STACPolicy=X-VLA2026.06 | 35 | 46 |