List your company

Adversarial Attack on StrongREJECT (Model Specific ASR)

11.7ASR (GPT-4o mini)Jailbroken
-0.4682.6915.859.009May 6, 2026

Evaluation results

MethodLinks
JailbrokenSetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.0511.73.33.3752520
FFASetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.0551.7093.328.383.3
Misrouter+FFASetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.0558.33.393.331.785
WildPromptSetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.051.71.71.791.721.720
MisrouterSetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.051.701.78.353.3
Direct InstructionSetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.050006.71.70
GCGSetting=Black-box transfer setting, Surrogate Source=Surrogate MoE models2026.0500001.71.7