Jailbreak Detection on JBB-Behaviors Benign
99.96AccuracySentra-Guard
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Sentra-GuardPrompt Types=100 safe but semantically close prompts.2025.10 | 99.96 | 99.98 | 99.94 | 99.96 | 0 |
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Sentra-GuardPrompt Types=100 safe but semantically close prompts.2025.10 | 99.96 | 99.98 | 99.94 | 99.96 | 0 |