Calibration
HLE → ARC-AGI-3
ARC-AGI-3 is estimated from HLE with a offset logistic curve fitted on 5 models measured on both: y = 0.0027 + (0.3016 − 0.0027) / (1 + exp(−200.00·(x − 0.5947))), R² = 1.00, cross-validated error 11.5 pp. It is used for 4 estimates.
| Estimated model | HLE | ARC-AGI-3 | Source |
|---|---|---|---|
| Agents-A1 | 47.6% | 0.3% | estimated ± 11.5 pp, low confidence |
| dots3-note Preview | 52.6% | 0.3% | estimated ± 11.5 pp, low confidence |
| GPT-5.4 Pro | 58.7% | 5.6% | estimated ± 11.5 pp, low confidence |
| GPT-5.5 Pro | 57.2% | 0.6% | estimated ± 11.5 pp, low confidence |