Calibration

MMLU-Pro (Arcee) → AA-Omniscience Accuracy

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

AA-Omniscience Accuracy is estimated from MMLU-Pro (Arcee) with a offset logistic curve fitted on 6 models measured on both: y = 0.2394 + (0.4655 − 0.2394) / (1 + exp(−168.45·(x − 0.8710))), R² = 0.97, cross-validated error 3.4 pp. It is used for 0 estimates.