Calibration

Toolathlon → APEX-Agents-AA

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

APEX-Agents-AA is estimated from Toolathlon with a Hill curve fitted on 14 models measured on both: y = 0.1378 + (1.2000 − 0.1378)·x^6.00 / (0.68115^6.00 + x^6.00), R² = 0.58, cross-validated error 8.0 pp. It is used for 0 estimates.