Hallucination Detection on CovidQA
91.7F1 ScoreHalluClean
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HalluCleanBackbone=GPT-3.5-turbo2025.11 | 91.7 | 92 | |
| DeepSeek-V32025.11 | 81.6 | 84 | |
| DeepSeek-R12025.11 | 68.4 | 75.5 | |
| GPT-4o-mini2025.11 | 53.2 | 67.5 | |
| GPT-3.5-turbo2025.11 | 9.5 | 52.5 | |
| Llama-3-70B2025.11 | 7.7 | 52 |