Trigger Inversion on Yahoo high poison rate
29Success RateCSO(discrete)
Evaluation Results
| Method | Links | |
|---|---|---|
| CSO(discrete)Model=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 29 | |
| CSO(discrete)Model=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 26 | |
| ALTModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 12 | |
| CSO(discrete)Model=Flan-t5-small, Target Trigger=mb(1 token), Label Type=dirty label, Poison Rate=high2026.06 | 10 | |
| MMModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=clean label, Poison Rate=high2026.06 | 10 | |
| MMModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=dirty label, Poison Rate=high2026.06 | 10 | |
| ALTModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 10 | |
| ALTModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=clean label, Poison Rate=high2026.06 | 10 | |
| ALTModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=dirty label, Poison Rate=high2026.06 | 10 | |
| CSO(discrete)Model=Flan-t5-small, Target Trigger=mb(1 token), Label Type=clean label, Poison Rate=high2026.06 | 9 | |
| UATModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 9 | |
| MMModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 8 | |
| UATModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 8 | |
| MMModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 5 | |
| GBDAModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 5 | |
| DBSModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=clean label, Poison Rate=high2026.06 | 3 | |
| DBSModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 2 | |
| DBSModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 2 | |
| PICCOLOModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 2 | |
| PICCOLOModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=dirty label, Poison Rate=high2026.06 | 2 | |
| DBSModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=dirty label, Poison Rate=high2026.06 | 1 | |
| GBDAModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=dirty label, Poison Rate=high2026.06 | 1 | |
| UATModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=clean label, Poison Rate=high2026.06 | 0 | |
| UATModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=dirty label, Poison Rate=high2026.06 | 0 | |
| GBDAModel=Flan-t5-small, Target Trigger=Tell me seriously(3 tokens), Label Type=clean label, Poison Rate=high2026.06 | 0 | |
| GBDAModel=Flan-t5-small, Target Trigger=mb(1 token), Label Type=clean label, Poison Rate=high2026.06 | 0 |