Safety Refusal on Safety Evaluation Prompts
62.5RefusalDirAbl
Evaluation Results
| Method | Links | |
|---|---|---|
| DirAblModel=Qwen2.5-1.5B, Steering Configuration=Best steering2026.01 | 62.5 | |
| DirAblModel=Llama-3.1-8B, Steering Configuration=Best steering2026.01 | 52.88 | |
| DirAblModel=Llama-3.2-3B, Steering Configuration=Best steering2026.01 | 24.04 | |
| DirAblModel=Llama-3.2-1B, Steering Configuration=Best steering2026.01 | 20.19 | |
| Selective SteeringModel=Llama-3.1-8B, Steering Configuration=Best steering2026.01 | 8.65 | |
| DirAblModel=gemma-2-9b, Steering Configuration=Best steering2026.01 | 7.69 | |
| DirAblModel=Qwen2.5-7B, Steering Configuration=Best steering2026.01 | 5.77 | |
| DirAblModel=Qwen2.5-3B, Steering Configuration=Best steering2026.01 | 1.92 | |
| DirAblModel=gemma-2-2b, Steering Configuration=Best steering2026.01 | 1.92 | |
| ActAddModel=Llama-3.1-8B, Steering Configuration=Best steering2026.01 | 0.96 | |
| SASModel=Llama-3.1-8B, Steering Configuration=Best steering2026.01 | 0.96 | |
| AASModel=Llama-3.1-8B, Steering Configuration=Best steering2026.01 | 0.96 | |
| ActAddModel=Llama-3.2-1B, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=Llama-3.2-1B, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=Llama-3.2-1B, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=Llama-3.2-1B, Steering Configuration=Best steering2026.01 | 0 | |
| ActAddModel=Llama-3.2-3B, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=Llama-3.2-3B, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=Llama-3.2-3B, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=Llama-3.2-3B, Steering Configuration=Best steering2026.01 | 0 | |
| ActAddModel=Qwen2.5-1.5B, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=Qwen2.5-1.5B, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=Qwen2.5-1.5B, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=Qwen2.5-1.5B, Steering Configuration=Best steering2026.01 | 0 | |
| ActAddModel=Qwen2.5-3B, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=Qwen2.5-3B, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=Qwen2.5-3B, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=Qwen2.5-3B, Steering Configuration=Best steering2026.01 | 0 | |
| ActAddModel=Qwen2.5-7B, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=Qwen2.5-7B, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=Qwen2.5-7B, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=Qwen2.5-7B, Steering Configuration=Best steering2026.01 | 0 | |
| ActAddModel=gemma-2-2b, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=gemma-2-2b, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=gemma-2-2b, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=gemma-2-2b, Steering Configuration=Best steering2026.01 | 0 | |
| ActAddModel=gemma-2-9b, Steering Configuration=Best steering2026.01 | 0 | |
| SASModel=gemma-2-9b, Steering Configuration=Best steering2026.01 | 0 | |
| AASModel=gemma-2-9b, Steering Configuration=Best steering2026.01 | 0 | |
| Selective SteeringModel=gemma-2-9b, Steering Configuration=Best steering2026.01 | 0 |