Jailbreak Detection on DrAttack using ASR and PGR
3ASRSelfDefend (Intent)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SelfDefend (Intent)Latency (Sec.)=2.38, Memory Overhead (MB)=13247.352026.04 | 3 | 20 | |
| GuardReasoner (Post)Latency (Sec.)=13.14, Memory Overhead (MB)=15317.422026.04 | 3 | 92 | |
| SelfGraderLatency (Sec.)=0.78, Memory Overhead (MB)=377.612026.04 | 3 | 17 | |
| GradientCuffLatency (Sec.)=19.26, Memory Overhead (MB)=1223.192026.04 | 4 | 10 | |
| Token HighlighterLatency (Sec.)=15.20, Memory Overhead (MB)=65536.662026.04 | 6 | 9 | |
| SelfDefend (Direct)Latency (Sec.)=0.80, Memory Overhead (MB)=13241.862026.04 | 6 | 54 | |
| WildGuard (Post)Latency (Sec.)=2.78, Memory Overhead (MB)=28098.392026.04 | 6 | 95 | |
| WildGuard (Pre)Latency (Sec.)=2.74, Memory Overhead (MB)=28014.702026.04 | 8 | 50 | |
| GuardReasoner (Pre)Latency (Sec.)=11.90, Memory Overhead (MB)=15317.412026.04 | 8 | 36 | |
| Llama Guard (Post)Latency (Sec.)=0.70, Memory Overhead (MB)=13783.602026.04 | 9 | 99 | |
| LLama-3-8B-Instruct (No Defense)Latency (Sec.)=0.862026.04 | 10 | — | |
| Perplexity FilterLatency (Sec.)=0.27, Memory Overhead (MB)=13279.462026.04 | 10 | 100 | |
| GradSafe2026.04 | 10 | 42 | |
| Prompt GuardLatency (Sec.)=16.87, Memory Overhead (MB)=1064.712026.04 | 10 | 94 | |
| Llama Guard (Pre)Latency (Sec.)=0.62, Memory Overhead (MB)=13697.272026.04 | 10 | 84 |