Prompt classification on XSTest
94.8F1 ScoreWildGuard
Evaluation Results
| Method | Links | |
|---|---|---|
| WildGuardModel Size=7B2026.01 | 94.8 | |
| YuFeng-XGuardModel Size=8B2026.01 | 94.4 | |
| PolyGuardModel Size=7B2026.01 | 92.2 | |
| YuFeng-XGuardModel Size=0.6B2026.01 | 91.6 | |
| Qwen3Guard-GenModel Size=8B, Evaluation protocol=strict2026.01 | 90.8 | |
| Qwen3Guard-GenModel Size=4B, Evaluation protocol=strict2026.01 | 89.9 | |
| GPT-OSS-SafeGuardModel Size=20B2026.01 | 89.9 | |
| Qwen3Guard-GenModel Size=8B, Evaluation protocol=loose2026.01 | 89.3 | |
| Llama3GuardModel Size=8B2026.01 | 88.4 | |
| Qwen3Guard-GenModel Size=4B, Evaluation protocol=loose2026.01 | 87.7 | |
| Qwen3Guard-GenModel Size=0.6B, Evaluation protocol=strict2026.01 | 85.7 | |
| NemotronReasoningModel Size=4B2026.01 | 84.8 | |
| Qwen3Guard-GenModel Size=0.6B, Evaluation protocol=loose2026.01 | 84.6 | |
| NemotronGuardV2Model Size=8B2026.01 | 84.1 | |
| Llama4GuardModel Size=12B2026.01 | 83.3 | |
| ShieldGemmaModel Size=9B2026.01 | 82.8 |