Failure Detection on Stack T Real-world
98AUCPRACC
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ACCPolicy=X-VLA2026.06 | 98 | 9 | |
| FAILPolicy=X-VLA2026.06 | 98 | 8 | |
| DiffPolicy=X-VLA2026.06 | 97 | 8 | |
| LLMDPolicy=X-VLA2026.06 | 97 | 8 | |
| ACEPolicy=X-VLA2026.06 | 96 | 9 | |
| FAILPolicy=π0.52026.06 | 96 | 10 | |
| ACCPolicy=π0.52026.06 | 95 | 12 | |
| LLMDPolicy=π0.52026.06 | 95 | 10 | |
| DiffPolicy=π0.52026.06 | 92 | 13 | |
| STACPolicy=π0.52026.06 | 87 | 17 | |
| ACEPolicy=π0.52026.06 | 82 | 24 | |
| STACPolicy=X-VLA2026.06 | 59 | 21 |