Span-level Hallucination Detection on HalluRAG (test)
80.56F1 ScoreLaplacian
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LaplacianModel=Mistral-7B, sliding-window=82026.02 | 80.56 | 89.2 | |
| Fourier-highModel=Mistral-7B, sliding-window=82026.02 | 78.83 | 88.72 | |
| Lookback-lensModel=Mistral-7B, sliding-window=82026.02 | 77.52 | 84.03 | |
| Wavelet-HighModel=Mistral-7B, sliding-window=82026.02 | 76.82 | 88.02 | |
| Wavelet-HighModel=LLaMA-13B, sliding-window=82026.02 | 74.17 | 84.38 | |
| LaplacianModel=LLaMA-13B, sliding-window=82026.02 | 72.17 | 82.64 | |
| Fourier-highModel=LLaMA-13B, sliding-window=82026.02 | 72.17 | 85.15 | |
| Lookback-lensModel=LLaMA-13B, sliding-window=82026.02 | 70.96 | 85.05 | |
| Fourier-highModel=LLaMA-7B, sliding-window=82026.02 | 68.66 | 81 | |
| Wavelet-HighModel=LLaMA-7B, sliding-window=82026.02 | 68.12 | 79.28 | |
| LaplacianModel=LLaMA-7B, sliding-window=82026.02 | 67.75 | 77.54 | |
| Lookback-lensModel=LLaMA-7B, sliding-window=82026.02 | 66.23 | 78.56 |