Calibration

CritPt → HLE w/o tools

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

HLE w/o tools is estimated from CritPt with a Hill curve fitted on 31 models measured on both: y = 0.0861 + (1.2000 − 0.0861)·x^0.49 / (0.82714^0.49 + x^0.49), R² = 0.79, cross-validated error 7.2 pp. It is used for 0 estimates.