Calibration
ZeroBench → SimpleVQA
SimpleVQA is estimated from ZeroBench with a Hill curve fitted on 5 models measured on both: y = 0.7439 + (0.0000 − 0.7439)·x^6.00 / (0.53170^6.00 + x^6.00), R² = 0.95, cross-validated error 2.5 pp. It is used for 3 estimates.
| Estimated model | ZeroBench | SimpleVQA | Source |
|---|---|---|---|
| Kimi K3 | 23.0% | 73.9% | estimated ± 2.5 pp, medium confidence |
| Seed 2.1 Pro | 18.0% | 74.3% | estimated ± 2.5 pp, low confidence |
| Seed 2.1 Turbo | 11.0% | 74.4% | estimated ± 2.5 pp, low confidence |