Backdoor Attack on Wiki topic evaluation requests (test)
93.2ASRBadDLM
Evaluation results
| Method | Links | |
|---|---|---|
| BadDLMMethod=BadDLM (Ours), Base Model=Dream-Instruct-7B2026.05 | 93.2 | |
| RL-basedMethod=RL-based, Base Model=Dream-Instruct-7B2026.05 | 67.5 | |
| VPIMethod=VPI, Base Model=Dream-Instruct-7B2026.05 | 54.1 | |
| SFT-basedMethod=SFT-based, Base Model=Dream-Instruct-7B2026.05 | 48.9 | |
| Benign (No Attack)Method=Benign (No Attack), Base Model=Dream-Instruct-7B2026.05 | 0 |