Token-level hallucination detection on RAGTruth
59.4AP (Token-level)
CORTEX
Updated 24d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.06 | 59.4 | 93.72 | |
2026.06 | 56.86 | 92.75 | |
2026.06 | 54.73 | 92.44 | |
2026.06 | 45.75 | 90.75 | |
2026.06 | 42.71 | 89.53 | |
2026.06 | 42.44 | 88.84 | |
2026.06 | 41.82 | 88.2 | |
2026.06 | 40.27 | 88.22 | |
2026.06 | 38.94 | 88.79 | |
2026.06 | 26.65 | 53.8 | |
2026.06 | 26.21 | 53.11 | |
2026.06 | 24.64 | 50.67 | |
2026.06 | 9.51 | 64.8 | |
2026.06 | 7.69 | 60.79 | |
2026.06 | 7.43 | 61.89 | |
2026.06 | 4.95 | 46.86 | |
2026.06 | 4.79 | 46.66 | |
2026.06 | 4.76 | 46.76 |