Targeted Advertisement Injection on Advertisement Injection
69Attack Success Rate (ARR)Poisoned-RM (Thought-Transfer)
Evaluation Results
| Method | Links | |
|---|---|---|
| Poisoned-RM (Thought-Transfer)Objective=Advertisement Injection2026.01 | 69 | |
| Qwen-14B (Base Model)Objective=Advertisement Injection2026.01 | 0 |