Evaluation Reliability on ToolBench (TB)
0.82Krippendorff's AlphaADARUBRIC-DA
Evaluation Results
| Method | Links | |
|---|---|---|
| ADARUBRIC-DA2026.03 | 0.82 | |
| ADARUBRIC-GM2026.03 | 0.8 | |
| ADARUBRIC-WM2026.03 | 0.79 | |
| Prometheus2026.03 | 0.68 | |
| GPT-4 Direct2026.03 | 0.66 | |
| G-EvalBackbone=GPT-4o2026.03 | 0.61 |