Specification Alignment on Specification Alignment Evaluation Set (full (1500))
76.4Safety Score
Align3
Updated 1mo ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2025.09 | 76.4 | 74.84 | 62.92 | |
2025.09 | 70.07 | 73.45 | 57.3 | |
2025.09 | 70 | 73.76 | 57.32 | |
2025.09 | 68.53 | 78.28 | 58.76 | |
2025.09 | 67.2 | 77.59 | 57.97 | |
2025.09 | 64.27 | 70.58 | 51.03 | |
2025.09 | 64.27 | 70.58 | 51.03 | |
2025.09 | 64.2 | 75.29 | 53.21 | |
2025.09 | 58.67 | 56.97 | 42.75 | |
2025.09 | 57.27 | 72.06 | 48.03 | |
2025.09 | 57.2 | 71.92 | 47.71 | |
2025.09 | 56.87 | 65.99 | 44.54 | |
2025.09 | 55.67 | 47.15 | 36.95 | |
2025.09 | 55.53 | 45.65 | 35.99 | |
2025.09 | 53.67 | 45.44 | 35.01 | |
2025.09 | 52.8 | 43.45 | 35.16 |