Jailbreak Detection on SB
98.33CORgpt-4o-mini
Evaluation Results
| Method | Links | |
|---|---|---|
| gpt-4o-miniVersion=2024-07-182026.02 | 98.33 | |
| gpt-4.1-miniVersion=2025-04-142026.02 | 98.33 | |
| gpt-5-miniVersion=2025-08-072026.02 | 98.33 | |
| AISABackbone=Mistral-7b-I2026.02 | 96.67 | |
| AISABackbone=Llama3.1-8b-I2026.02 | 96.67 | |
| AISABackbone=Llama2-13b-I2026.02 | 96.67 | |
| AISABackbone=Qwen3-8b-I2026.02 | 95.83 | |
| GradSafe2026.02 | 95 | |
| AISABackbone=GPT-OSS-20b-I2026.02 | 94.17 | |
| Jailbreak-Classifier2026.02 | 92.5 | |
| NemoGuard-JailbreakDetect2026.02 | 79.17 | |
| SPDetector2026.02 | 76.67 | |
| Llama-Prompt-Guard-2Parameters=86M2026.02 | 8.33 |