Priority Scoring (Description) on CyberThreat-Eval
39.82Precision (Accepted)o3-mini
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| o3-miniFine-tuned=false2026.03 | 39.82 | 90.91 | 55 | 0.68 | 5,178.38 | 450,093 | |
| GPT-4oFine-tuned=false2026.03 | 35.9 | 98.99 | 56.6 | 0.505 | 708.23 | 430,124 | |
| GPT-4o-mini (FT)Fine-tuned=true2026.03 | 34.77 | 97.98 | 59 | 0.52 | 4,424.71 | 555,547 | |
| GPT-4o (FT)Fine-tuned=true2026.03 | 33.92 | 97.98 | 60 | 0.47 | 3,939.35 | 453,280 |