Calibration

ZeroBench → MathVision

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

MathVision is estimated from ZeroBench with a Hill curve fitted on 5 models measured on both: y = 0.8941 + (1.2000 − 0.8941)·x^6.00 / (0.30847^6.00 + x^6.00), R² = 0.59, cross-validated error 2.7 pp. It is used for 0 estimates.