Hallucination Detection on HaluEval English - Base (test)
98.5AUROCMultiHaluDet
Evaluation Results
| Method | Links | |
|---|---|---|
| MultiHaluDetBackbone=LLaMA2-7B2026.05 | 98.5 | |
| MultiHaluDetBackbone=Mistral-7B2026.05 | 98.4 | |
| Best BaselineBackbone=Mistral-7B2026.05 | 95.4 | |
| Best BaselineBackbone=LLaMA2-7B2026.05 | 92.8 |