List your company

Jailbreak Attack on AdvBench (StrongREJECT score)

98StrongREJECT ScoreDictAttack
6.4830.245477.76Mar 31, 2025

Evaluation results

MethodLinks
DictAttackModel=gemini-2.5-pro2025.0398
AutoDAN-TurboModel=gemini-2.5-pro2025.0387
TAPModel=gemini-2.5-pro2025.0323
PAIRModel=gemini-2.5-pro2025.0310