Jailbreak Defense Efficiency on HarmBench
0Additional TokensGUARD-SLM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GUARD-SLM2026.03 | 0 | 0.43 | |
| RobustAlign2026.03 | 1 | 12.45 | |
| SmoothLLM2026.03 | 1 | 38.07 | |
| Aligner2026.03 | 42 | 4.9 | |
| SelfReminder2026.03 | 44 | 1.12 | |
| SelfEval2026.03 | 51 | 8.24 | |
| GoalPrior2026.03 | 202 | 3.18 | |
| ICD2026.03 | 805 | 1.13 |