Calibration

ERQA → MMMU-Pro

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

MMMU-Pro is estimated from ERQA with a offset logistic curve fitted on 10 models measured on both: y = 0.7603 + (0.8141 − 0.7603) / (1 + exp(−200.00·(x − 0.6402))), R² = 0.76, cross-validated error 1.7 pp. It is used for 0 estimates.