Jailbreak Evaluation on JailBreak R1
1.3Attack Success Rate (ASR)SInternal
Evaluation Results
| Method | Links | |
|---|---|---|
| SInternalModel Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 1.3 | |
| STAR-1Model Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 5.1 | |
| SInternalModel Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 6.1 | |
| STAR-1Model Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 9.3 | |
| SInternalModel Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 11.8 | |
| STAR-1Model Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 15.3 | |
| SafeChainModel Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 32.9 | |
| SafeChainModel Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 39 | |
| BaseModel Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 41.5 | |
| SafeChainModel Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 46.7 | |
| BaseModel Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 49.5 | |
| BaseModel Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 54 |