Phishing Worm on Single-Agent Evaluation Set
100R@5CEM Attack
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| CEM AttackModel=GPT-4o2026.01 | 100 | 77 | — | — | |
| fusion attackModel=GPT-4o2026.01 | 100 | 81 | — | — | |
| CEM AttackModel=GPT-4o-mini2026.01 | 100 | 77 | — | — | |
| fusion attackModel=GPT-4o-mini2026.01 | 100 | 81 | — | — | |
| Query+Model=GPT-4o-mini2026.01 | 63 | 70 | — | — | |
| Query+Model=GPT-4o2026.01 | 56 | 70 | — | — |