Calibration

CWE-bench v1 → AA Agentic Index

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

AA Agentic Index is estimated from CWE-bench v1 with a inverse Michaelis–Menten curve fitted on 5 models measured on both: y = 1.24741·(x − 0.0000) / (0.0000 + 2.0000 − x), R² = 0.59, cross-validated error 13.3 pp. It is used for 0 estimates.