Safety Evaluation on StrongReject (Harmful Score)
0Harmful ScoreBase
Evaluation results
| Method | Links | |
|---|---|---|
| BaseModel Backbone=DeepSeek-R1-Distill-Qwen-7B, RL Training (GRPO)=true2026.05 | 0 | |
| STAR-1Model Backbone=DeepSeek-R1-Distill-Llama-8B, RL Training (GRPO)=true2026.05 | 0 | |
| SInternalModel Backbone=DeepSeek-R1-Distill-Llama-8B, RL Training (GRPO)=true2026.05 | 0 | |
| BaseModel Backbone=DeepSeek-R1-Distill-Qwen-14B, RL Training (GRPO)=true2026.05 | 0 | |
| SafeChainModel Backbone=DeepSeek-R1-Distill-Qwen-14B, RL Training (GRPO)=true2026.05 | 0 | |
| STAR-1Model Backbone=DeepSeek-R1-Distill-Qwen-14B, RL Training (GRPO)=true2026.05 | 0 | |
| SInternalModel Backbone=DeepSeek-R1-Distill-Qwen-14B, RL Training (GRPO)=true2026.05 | 0 | |
| SInternalModel Backbone=DeepSeek-R1-Distill-Qwen-7B, RL Training (GRPO)=true2026.05 | 0.3 | |
| SafeChainModel Backbone=DeepSeek-R1-Distill-Llama-8B, RL Training (GRPO)=true2026.05 | 0.3 | |
| SafeChainModel Backbone=DeepSeek-R1-Distill-Qwen-7B, RL Training (GRPO)=true2026.05 | 1 | |
| STAR-1Model Backbone=DeepSeek-R1-Distill-Qwen-7B, RL Training (GRPO)=true2026.05 | 1 | |
| BaseModel Backbone=DeepSeek-R1-Distill-Llama-8B, RL Training (GRPO)=true2026.05 | 3.2 |