Factual Accuracy Estimation on Real-data Factual GPT-5.4 responses
64.39Original AccuracyLogistic Regression
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Logistic RegressionQuestion Type=Factual, Data=2842, Verification Budget=3%2026.06 | 64.39 | 58.72 | -5.68 | 85 | |
| Multilayer PerceptronQuestion Type=Factual, Data=2842, Verification Budget=3%2026.06 | 64.39 | 59.01 | -5.38 | 85 | |
| Random ForestQuestion Type=Factual, Data=2842, Verification Budget=3%2026.06 | 64.39 | 63.4 | -0.99 | 85 | |
| Label PropagationQuestion Type=Factual, Data=2842, Verification Budget=3%2026.06 | 64.39 | 54.87 | -9.52 | 85 | |
| nnPUQuestion Type=Factual, Data=2842, Verification Budget=3%2026.06 | 64.39 | 64.39 | 0 | 85 | |
| AURAQuestion Type=Factual, Data=2842, Verification Budget=3%2026.06 | 64.39 | 67.66 | 3.27 | 95.6 |