Calibration

AA-HLE → AA-GPQA Diamond

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

AA-GPQA Diamond is estimated from AA-HLE with a Michaelis–Menten + offset curve fitted on 179 models measured on both: y = 0.2375 + 0.7958·x / (0.07189 + x), R² = 0.80, cross-validated error 7.0 pp. It is used for 0 estimates.