Safety Evaluation on XSTest Toxic
95SafetySTL
Evaluation Results
| Method | Links | |
|---|---|---|
| STLBase Model=Qwen2.5-7B2026.02 | 95 | |
| STL-augBase Model=Qwen2.5-7B2026.02 | 95 | |
| SCANSBase Model=Qwen2.5-7B2026.02 | 94 | |
| DCRBase Model=Qwen2.5-7B2026.02 | 94 | |
| SurgicalBase Model=Qwen2.5-7B2026.02 | 93 | |
| STLBase Model=LLaMA-3-8B2026.02 | 93 | |
| STL-augBase Model=LLaMA-3-8B2026.02 | 91 | |
| SurgicalBase Model=LLaMA-3-8B2026.02 | 91 | |
| DCRBase Model=LLaMA-3-8B2026.02 | 91 | |
| SCANSBase Model=LLaMA-3-8B2026.02 | 88 | |
| DCRBase Model=Qwen2.5-1.5B2026.02 | 81 | |
| SurgicalBase Model=Qwen2.5-1.5B2026.02 | 78 | |
| STL-augBase Model=Qwen2.5-1.5B2026.02 | 77 | |
| STLBase Model=Qwen2.5-1.5B2026.02 | 72 | |
| SCANSBase Model=Qwen2.5-1.5B2026.02 | 65 |