Sponge Attack Detection on GCG-DOS
99.85F1 ScoreSHIELD
Evaluation Results
| Method | Links | |
|---|---|---|
| SHIELDTarget Model=LLaMA22026.01 | 99.85 | |
| Harm-filterTarget Model=LLaMA22026.01 | 96.86 | |
| Perplexity-filterTarget Model=LLaMA22026.01 | 96.07 |
| Method | Links | |
|---|---|---|
| SHIELDTarget Model=LLaMA22026.01 | 99.85 | |
| Harm-filterTarget Model=LLaMA22026.01 | 96.86 | |
| Perplexity-filterTarget Model=LLaMA22026.01 | 96.07 |