Priority Scoring (Article) on CyberThreat-Eval
0.3802Precision (Accepted)o3-mini
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| o3-miniFine-tuned=false2026.03 | 0.3802 | 0.9293 | 0.58 | 0.55 | 6,048.47 | 1,390,052 | |
| GPT-4oFine-tuned=false2026.03 | 0.3037 | 1 | 0.566 | 0.505 | 785.64 | 1,438,448 | |
| GPT-4o-mini (FT)Fine-tuned=true2026.03 | 0.2964 | 1 | 0.6 | 0.45 | 7,283.13 | 1,891,248 | |
| GPT-4o (FT)Fine-tuned=true2026.03 | 0.2717 | 0.9798 | 0.59 | 0.5 | 6,034.71 | 1,522,969 |