Calibration

DeepSearchQA → Gert Labs

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

Gert Labs is estimated from DeepSearchQA with a offset logistic curve fitted on 8 models measured on both: y = 0.3836 + (0.5900 − 0.3836) / (1 + exp(−200.00·(x − 0.6862))), R² = 0.44, cross-validated error 11.2 pp. It is used for 0 estimates.