Calibration

MedXpertQA (Text) → GPQA Diamond

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

GPQA Diamond is estimated from MedXpertQA (Text) with a offset logistic curve fitted on 5 models measured on both: y = 0.8701 + (0.9437 − 0.8701) / (1 + exp(−28.55·(x − 0.5501))), R² = 1.00, cross-validated error 0.9 pp. It is used for 0 estimates.