List your company

Adversarial Robustness on WildJailbreak 2k queries (test)

498Number of Explicit RefusalsAUTOSKILL
110.08210.79311.5412.21Apr 19, 2026

Evaluation results

MethodLinks
AUTOSKILLBudget=1002026.0449837
RandomBudget=1002026.04461—
AUTOSKILLBudget=902026.0445215
AUTOSKILLBudget=802026.0443735
RandomBudget=902026.04437—
AUTOSKILLBudget=702026.0440223
RandomBudget=802026.04402—
AUTOSKILLBudget=602026.0438812
RandomBudget=702026.04379—
RandomBudget=602026.04376—
AUTOSKILLBudget=502026.0414116
RandomBudget=502026.04125—