Calibration

FrontierCode 1.1 Main → AA-SciCode

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

AA-SciCode is estimated from FrontierCode 1.1 Main with a linear curve fitted on 11 models measured on both: y = 0.3346·x + 0.4182, R² = 0.40, cross-validated error 3.8 pp. It is used for 2 estimates.

Estimated model	FrontierCode 1.1 Main	AA-SciCode	Source
SWE-1.7	42.3%	56.0%	estimated ± 3.8 pp, medium confidence
SWE-2	50.0%	58.5%	estimated ± 3.8 pp, medium confidence