Failure Detection on Cups Real-world
98AUCPR (PR)STAC
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| STACPolicy=X-VLA2026.06 | 98 | 42 | |
| ACCPolicy=X-VLA2026.06 | 97 | 28 | |
| FAILPolicy=X-VLA2026.06 | 93 | 28 | |
| DiffPolicy=X-VLA2026.06 | 91 | 27 | |
| ACEPolicy=X-VLA2026.06 | 89 | 28 | |
| LLMDPolicy=X-VLA2026.06 | 78 | 33 | |
| FAILPolicy=π0.52026.06 | 63 | 49 | |
| STACPolicy=π0.52026.06 | 61 | 52 | |
| ACCPolicy=π0.52026.06 | 60 | 51 | |
| LLMDPolicy=π0.52026.06 | 58 | 52 | |
| DiffPolicy=π0.52026.06 | 51 | 57 | |
| ACEPolicy=π0.52026.06 | 41 | 61 |