List your company

Jailbreaking on AdvBench Puzzler prompts llama-3.1-8b

100ASRPuzzler
-4235077Mar 31, 2025

Evaluation results

MethodLinks
PuzzlerDefense=None, Eval Size=202025.03100
EnumAttackDefense=None, Eval Size=20, Payload Template=Puzzler2025.03100
EnumAttackDefense=JBShield, Eval Size=20, Payload Template=Puzzler2025.03100
PuzzlerDefense=JBShield, Eval Size=202025.030