ResearchBenchmarksAdversarial Attack on Claude Opus 4.1 (test)Follow46.7Bin-ASRPLAGUE7.90817.97928.0538.121Oct 20, 2025Evaluation ResultsMethodMethodLinksBin-ASRSREPLAGUEInteraction budget=equ...Interaction budget=equal budget (six queries per behavior), Configuration=Best2025.1046.767.3FITDAttacker model=DeepSee...Attacker model=DeepSeek-R1, Interaction budget=six queries per behavior2025.1010.556X-TeamingAttacker model=DeepSee...Attacker model=DeepSeek-R1, Interaction budget=six queries per behavior2025.109.422.8