Calibration

FrontierSWE v2 → PostTrainBench v1.1

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

PostTrainBench v1.1 is estimated from FrontierSWE v2 with a linear curve fitted on 7 models measured on both: y = 0.3388·x + 0.2343, R² = 0.66, cross-validated error 4.3 pp. It is used for 2 estimates.

Estimated model	FrontierSWE v2	PostTrainBench v1.1	Source
Claude Sonnet 5.5	61.9%	44.4%	estimated ± 4.3 pp, medium confidence
Grok 4.7	29.5%	33.4%	estimated ± 4.3 pp, medium confidence