Calibration
HLE → HLE w/o tools
HLE w/o tools is estimated from HLE with a linear curve fitted on 30 models measured on both: y = 0.8023·x + 0.0065, R² = 0.83, cross-validated error 5.3 pp. It is used for 3 estimates.
| Estimated model | HLE | HLE w/o tools | Source |
|---|---|---|---|
| Agents-A1 | 47.6% | 38.8% | estimated ± 5.3 pp, medium confidence |
| Apodex 1.1 | 56.1% | 45.7% | estimated ± 5.3 pp, medium confidence |
| dots3-note Preview | 52.6% | 42.9% | estimated ± 5.3 pp, medium confidence |