Calibration
Toolathlon-Verified → AA Agentic Index
AA Agentic Index is estimated from Toolathlon-Verified with a linear curve fitted on 7 models measured on both: y = 1.3035·x + -0.4596, R² = 0.93, cross-validated error 3.3 pp. It is used for 8 estimates.
| Estimated model | Toolathlon-Verified | AA Agentic Index | Source |
|---|---|---|---|
| dots3-note Preview | 55.6% | 26.5% | estimated ± 3.3 pp, medium confidence |
| GLM-5.3-Flash | 78.4% | 56.2% | estimated ± 3.3 pp, medium confidence |
| Hy4 preview | 74.1% | 50.6% | estimated ± 3.3 pp, medium confidence |
| Laguna S 2.1 | 49.7% | 18.8% | estimated ± 3.3 pp, low confidence |
| Ornith-1.5-35B-A3B | 48.7% | 17.5% | estimated ± 3.3 pp, low confidence |
| Ornith-1.5-397B | 71.2% | 46.8% | estimated ± 3.3 pp, medium confidence |
| Ornith-1.5-9B | 41.2% | 7.7% | estimated ± 3.3 pp, low confidence |
| Qwen3.8 Max | 72.5% | 48.5% | estimated ± 3.3 pp, medium confidence |