List your company

Adversarial Attack on OR-Bench unsafe inputs

88ASRGranite
8.9629.485070.52Jan 27, 2025

Evaluation results

MethodLinks
GraniteAttack Type=White-box, Input Setting=Unsafe inputs2025.0188
Qwen 2.5Attack Type=White-box, Input Setting=Unsafe inputs2025.0188
Gemma 1Attack Type=White-box, Input Setting=Unsafe inputs2025.0167
Llama 2Attack Type=White-box, Input Setting=Unsafe inputs2025.0112