Sponge Attack Detection on RL-GOAL
99.6F1 ScoreSHIELD
Evaluation Results
| Method | Links | |
|---|---|---|
| SHIELDTarget Model=LLaMA22026.01 | 99.6 | |
| Perplexity-filterTarget Model=LLaMA22026.01 | 99.19 | |
| Harm-filterTarget Model=LLaMA22026.01 | 93.71 |
| Method | Links | |
|---|---|---|
| SHIELDTarget Model=LLaMA22026.01 | 99.6 | |
| Perplexity-filterTarget Model=LLaMA22026.01 | 99.19 | |
| Harm-filterTarget Model=LLaMA22026.01 | 93.71 |