List your company

Jailbreak Defense on Adversarial Jailbreak Attacks (4 Specific Vectors)

100Safety Score (Cipher)Claude Opus 4.5
92.678494.579296.4898.3808Apr 24, 2026

Evaluation results

MethodLinks
Claude Opus 4.5Defense Pipeline=Graph-augmented pedagogical pipeline2026.0410094.3791.5594.37
Gemini 2.5 Flash-LiteDefense Pipeline=Graph-augmented pedagogical pipeline2026.0494.3789.4492.2591.55
Gemini 2.5 FlashDefense Pipeline=Graph-augmented pedagogical pipeline2026.0493.6692.9692.9638.03
GPT-5Defense Pipeline=Graph-augmented pedagogical pipeline2026.0493.6691.5590.8592.25
Gemini 2.5 ProDefense Pipeline=Graph-augmented pedagogical pipeline2026.0492.9692.9695.0789.44