Jailbreak Detection on JailbreakTracer Corpus
98.88AccuracySentra-Guard
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Sentra-GuardPrompt Types=Synthetic + real-world toxic prompts.2025.10 | 98.88 | 99.91 | 99.84 | 99.87 | 1.2 |
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Sentra-GuardPrompt Types=Synthetic + real-world toxic prompts.2025.10 | 98.88 | 99.91 | 99.84 | 99.87 | 1.2 |