LRM-specific Jailbreak Evaluation on HCoT
100ASRBase
Evaluation Results
| Method | Links | |
|---|---|---|
| BaseModel Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 100 | |
| SafeChainModel Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 100 | |
| BaseModel Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 100 | |
| SafeChainModel Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 100 | |
| STAR-1Model Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 100 | |
| BaseModel Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 100 | |
| SafeChainModel Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 100 | |
| STAR-1Model Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 100 | |
| STAR-1Model Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 98 | |
| SInternalModel Architecture=DeepSeek-R1-Distill-Llama-8B2026.05 | 94 | |
| SInternalModel Architecture=DeepSeek-R1-Distill-Qwen-7B2026.05 | 92 | |
| SInternalModel Architecture=DeepSeek-R1-Distill-Qwen-14B2026.05 | 90 |