benchgap
Calibration

AA-GPQA Diamond → ARC-AGI-1

ARC-AGI-1 is estimated from AA-GPQA Diamond with a Hill curve fitted on 28 models measured on both: y = 0.0000 + (1.2000 − 0.0000)·x^6.00 / (0.76802^6.00 + x^6.00), R² = 0.48, cross-validated error 8.6 pp. It is used for 0 estimates.