List your company

Jailbreak Mitigation on GCG

0GCG ASRGoal
-2.12812.23626.640.964Sep 30, 2025Nov 2, 2025Dec 5, 2025Jan 7, 2026Feb 9, 2026Mar 14, 2026Apr 17, 2026

Evaluation results

MethodLinks
GoalModel=Mistral-7B2026.040—————
ASGUARD2025.09196.759.568.37645
SFTData Mixture=30/702025.09266.12.2767.313.813.4
RepBend2025.09594.350.468.173.341.7
CBModel=Mistral-7B2026.0414—————
Llama-3.1-8B-InstructBackbone=Llama-3.1, Parameters=8B, Mode=Instruct2025.091588.528.968.2——
BeamModel=Mistral-7B2026.0417.5—————
BaseModel=Mistral-7B2026.0453.2—————