List your company

Jailbreak Detection on FQ-PH

86.54ACCJailbreak-Classifier
51.200860.375469.5578.7246Feb 14, 2026

Evaluation results

MethodLinks
Jailbreak-Classifier2026.0286.54
AISABackbone=Llama3.1-8b-I2026.0278.85
AISABackbone=Qwen3-8b-I2026.0278.21
AISABackbone=Mistral-7b-I2026.0277.56
AISABackbone=Llama2-13b-I2026.0277.56
NemoGuard-JailbreakDetect2026.0276.28
gpt-5-miniVersion=2025-08-072026.0275.64
gpt-4.1-miniVersion=2025-04-142026.0274.36
GradSafe2026.0273.72
gpt-4o-miniVersion=2024-07-182026.0272.44
AISABackbone=GPT-OSS-20b-I2026.0268.59
SPDetector2026.0263.46
Llama-Prompt-Guard-2Parameters=86M2026.0252.56