Input prompt classification on GSM8K
100AccuracyLlama-Guard-3-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama-Guard-3-8Bsize=8B2026.05 | 100 | |
| Qwen3Guard-8Bsize=8B2026.05 | 100 | |
| EvoSafety (Guard)-3Bsize=3B2026.05 | 100 | |
| EvoSafety+ (Guard)-3Bsize=3B, variant=plus2026.05 | 100 | |
| WildGuard-7Bsize=7B2026.05 | 99.92 |