Calibration

V* → ERQA

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

ERQA is estimated from V* with a Hill curve fitted on 5 models measured on both: y = 0.0000 + (0.7145 − 0.0000)·x^6.00 / (0.59205^6.00 + x^6.00), R² = 0.84, cross-validated error 5.6 pp. It is used for 1 estimate.

Estimated model	V*	ERQA	Source
Step 3.7 Flash	95.3%	67.6%	estimated ± 5.6 pp, low confidence