Safety Evaluation on Attack Benchmarks Response
82.3SEvalAYuFeng-XGuard
Evaluation results
| Method | Links | ||
|---|---|---|---|
| YuFeng-XGuardModel Scale=8B2026.01 | 82.3 | 82.3 | |
| GPT-OSS-SafeGuardModel Scale=20B2026.01 | 81.5 | 81.5 | |
| YuFeng-XGuardModel Scale=0.6B2026.01 | 79.6 | 79.6 | |
| NemotronReasoningModel Scale=4B2026.01 | 72.6 | 72.6 | |
| Qwen3Guard-GenModel Scale=8B, Evaluation Policy=strict2026.01 | 72.3 | 72.3 | |
| Qwen3Guard-GenModel Scale=4B, Evaluation Policy=strict2026.01 | 71.9 | 71.9 | |
| Qwen3Guard-GenModel Scale=0.6B, Evaluation Policy=strict2026.01 | 69.1 | 69.1 | |
| PolyGuardModel Scale=7B2026.01 | 67.9 | 67.9 | |
| NemotronGuardV2Model Scale=8B2026.01 | 66.7 | 66.7 | |
| Qwen3Guard-GenModel Scale=4B, Evaluation Policy=loose2026.01 | 62.9 | 62.9 | |
| Qwen3Guard-GenModel Scale=0.6B, Evaluation Policy=loose2026.01 | 62.7 | 62.7 | |
| Qwen3Guard-GenModel Scale=8B, Evaluation Policy=loose2026.01 | 62.1 | 62.1 | |
| Llama3GuardModel Scale=8B2026.01 | 56.6 | 56.6 | |
| Llama4GuardModel Scale=12B2026.01 | 55.8 | 55.8 | |
| WildGuardModel Scale=7B2026.01 | 54.6 | 54.6 | |
| ShieldGemmaModel Scale=9B2026.01 | 27.7 | 27.7 |