Safety Moderation on StrongReject
100F1 ScoreYuFeng-XGuard
Evaluation Results
| Method | Links | |
|---|---|---|
| YuFeng-XGuardSize=8B, Task=s-cls2026.05 | 100 | |
| GLiGuard OmniSize=209M, Task=s-cls+IE2026.05 | 99.7 | |
| WildGuardSize=7B, Task=s-cls2026.05 | 99.5 | |
| NemotronGuardV2Size=8B, Task=s-cls2026.05 | 99.5 | |
| GPT-OSS-SafeGuardSize=20B, Task=s-cls2026.05 | 99.2 | |
| Llama3GuardSize=8B, Task=s-cls2026.05 | 98.5 | |
| GLiGuard uni-encSize=147M, Task=s-cls+IE2026.05 | 98.5 | |
| GLiGuard bi-encSize=145M, Task=s-cls+IE2026.05 | 97.7 | |
| GLiNER2 Multi v1Size=209M, Task=cls+IE2026.05 | 95.7 | |
| Llama4GuardSize=12B, Task=s-cls2026.05 | 95.3 | |
| GLiClass Base v1.0Size=150M, Task=cls2026.05 | 91.3 | |
| ShieldGemmaSize=9B, Task=s-cls2026.05 | 90 | |
| Longformer-harmful-roSize=149M, Task=s-cls2026.05 | 81.7 | |
| PromptGuard 2*Size=86M, Task=s-cls2026.05 | 16.4 | |
| DeBERTa-v3-base-PISize=184M, Task=p-inj2026.05 | 0.6 |