List your company

Out-of-distribution (OOD) Harmful Content Detection on JailbreakBench

0.962AUROC (Alpaca)w_opt
0.91390.937950.9620.98605Apr 20, 2026

Evaluation results

MethodLinks
w_optstrategy=Optimised discriminative direction, aggregation=Mean across 12 models2026.040.9620.9750.962