Content Root Cause Analysis on CyberThreat-Eval
4.733Relevanceo3-mini
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| o3-miniFine-tuned=false2026.03 | 4.733 | 4.627 | 4.576 | 4.818 | 4.8 | 4.418 | |
| GPT-4o (FT)Fine-tuned=true2026.03 | 4.533 | 4.371 | 4.286 | 4.629 | 4.586 | 4.323 | |
| GPT-4o-mini (FT)Fine-tuned=true2026.03 | 3.808 | 3.819 | 3.441 | 4.163 | 3.988 | 3.667 | |
| GPT-4oFine-tuned=false2026.03 | 3.686 | 3.458 | 3.362 | 3.932 | 3.753 | 3.612 |