Factual Accuracy Estimation on Real-data Factual Qwen responses
55.94Original AccuracyLogistic Regression
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Logistic RegressionQuestion Type=Factual, Data=9333, Verification Budget=3%2026.06 | 55.94 | 56.15 | 0.21 | 280 | |
| Multilayer PerceptronQuestion Type=Factual, Data=9333, Verification Budget=3%2026.06 | 55.94 | 56.51 | 0.57 | 280 | |
| Random ForestQuestion Type=Factual, Data=9333, Verification Budget=3%2026.06 | 55.94 | 60.15 | 4.21 | 280 | |
| Label PropagationQuestion Type=Factual, Data=9333, Verification Budget=3%2026.06 | 55.94 | 51.89 | -4.05 | 280 | |
| nnPUQuestion Type=Factual, Data=9333, Verification Budget=3%2026.06 | 55.94 | 59.23 | 3.29 | 280 | |
| AURAQuestion Type=Factual, Data=9333, Verification Budget=3%2026.06 | 55.94 | 80.31 | 24.37 | 292.8 |