List your company

ResearchFormerly WizWand

Directional Jailbreak Detection

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

WildJailbreak (Safe) × StrongREJECT (Harmful)

98.1AUROC (Mono)

3

Jun 25, 2026

WildJailbreak (Safe) × HarmBench (Harmful)

1AUROC (mono)

3

Jun 25, 2026

WildJailbreak (Safe) × AdvBench (Harmful)

1AUROC (mono)

3

Jun 25, 2026

UltraChat (Safe) × StrongREJECT (Harmful)

0.797AUROC (mono)

3

Jun 25, 2026

UltraChat (Safe) × HarmBench (Harmful)

0.956AUROC (mono)

3

Jun 25, 2026

UltraChat (Safe) × AdvBench (Harmful)

0.971AUROC (mono)

3

Jun 25, 2026