benchgap
Calibration

BioBasicsBench → HealthBench Professional

HealthBench Professional is estimated from BioBasicsBench with a Michaelis–Menten curve fitted on 5 models measured on both: y = 0.8554·x / (0.20223 + x), R² = 0.70, cross-validated error 6.1 pp. It is used for 0 estimates.

Fitted on

ModelBioBasicsBenchHealthBench Professional
GPT-5.433.2%48.1%
GPT-5.6 Sol45.6%60.5%
GPT-6.1 Sol62.6%64.2%
GPT-6 Astra67.0%64.7%
GPT-6 Luna37.6%60.8%