Long-form generation factuality and uncertainty estimation on WildHallu (test)
0.86Factuality Score
LOGU-DPO
Updated 4d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2024.10 | 0.86 | 52.8 | 2.68 | |
2024.10 | 0.844 | 61.8 | 3.49 | |
2024.10 | 0.84 | 48.1 | 2.33 | |
2024.10 | 0.802 | 48.6 | 3.44 | |
2024.10 | 0.792 | 51.1 | 5.73 | |
2024.10 | 0.789 | 44.6 | 5.37 | |
2024.10 | 0.787 | 45.8 | 3.29 | |
2024.10 | 0.77 | 28.7 | 5.13 | |
2024.10 | 0.754 | 48.6 | 5.2 | |
2024.10 | 0.744 | - | 6.24 | |
2024.10 | 0.734 | 51.6 | 9.8 | |
2024.10 | 0.726 | 57.3 | 10.1 | |
2024.10 | 0.724 | 50.6 | 8.29 | |
2024.10 | 0.715 | - | 8.31 |