Calibration

VideoMMMU → MLVU (M-Avg)

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

MLVU (M-Avg) is estimated from VideoMMMU with a offset logistic curve fitted on 9 models measured on both: y = 0.8523 + (1.2000 − 0.8523) / (1 + exp(−200.00·(x − 0.8958))), R² = 0.43, cross-validated error 2.9 pp. It is used for 0 estimates.