ResearchBenchmarksJailbreak Detection on HILL jailbreak attacks (test)Follow98.03AUROCT3+GMM97.82297.87697.9397.984Feb 4, 2026Evaluation ResultsMethodMethodLinksAUROCFPR@95AUPRCF1 ScoreT3+GMMDensity Estimation Alg...Density Estimation Algorithm=Gaussian Mixture Model (GMM)2026.0298.034.3599.5498.81T3+OCSVMDensity Estimation Alg...Density Estimation Algorithm=One-Class SVM (OCSVM)2026.0297.834.3599.2698.61