Calibration

ResearchClawBench → BrowseComp

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

BrowseComp is estimated from ResearchClawBench with a offset logistic curve fitted on 9 models measured on both: y = 0.2722 + (0.8166 − 0.2722) / (1 + exp(−200.00·(x − 0.1382))), R² = 0.72, cross-validated error 7.1 pp. It is used for 1 estimate.

Estimated model	ResearchClawBench	BrowseComp	Source
Grok 4.1	13.5%	46.0%	estimated ± 7.1 pp, low confidence