Out-of-distribution (OOD) Harmful Content Detection on JailbreakBench
0.962AUROC (Alpaca)w_opt
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| w_optstrategy=Optimised discriminative direction, aggregation=Mean across 12 models2026.04 | 0.962 | 0.975 | 0.962 |
| Method | Links | |||
|---|---|---|---|---|
| w_optstrategy=Optimised discriminative direction, aggregation=Mean across 12 models2026.04 | 0.962 | 0.975 | 0.962 |