Calibration

AA Harvey LAB → HLE w/ tools

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

HLE w/ tools is estimated from AA Harvey LAB with a Hill curve fitted on 5 models measured on both: y = 0.0000 + (1.2000 − 0.0000)·x^6.00 / (0.90618^6.00 + x^6.00), R² = 0.89, cross-validated error 5.7 pp. It is used for 2 estimates.

Estimated model	AA Harvey LAB	HLE w/ tools	Source
Claude Fable 5	93.6%	65.8%	estimated ± 5.7 pp, low confidence
Gemini 3.7 Flash	90.7%	60.2%	estimated ± 5.7 pp, low confidence