Step-level Correctness Prediction on ToolBench
0.8129AUROC
PAIR
Updated 15d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.05 | 0.8129 | 22.36 | |
2026.05 | 0.7821 | 24.76 | |
2026.05 | 0.7765 | 19.61 | |
2026.05 | 0.7654 | 25.62 | |
2026.05 | 0.7598 | 26.83 | |
2026.05 | 0.7561 | 23.29 | |
2026.05 | 0.7372 | 28.71 | |
2026.05 | 0.7295 | 16.85 | |
2026.05 | 0.6847 | 19.98 | |
2026.05 | 0.5953 | - |