MultiJail
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Jailbreak Attack Success Rate | MultiJail | ASR (EN)13.02 | 18 | |
| Multilingual Safety | MultiJail Out-of-Distribution Languages (test) | Safety Violation Rate (KO)0.63 | 10 | |
| Multilingual Safety | MultiJail In-Distribution Languages (test) | Safety Score (EN)33.65 | 10 | |
| Safety Evaluation | MultiJail-AR (test) | Safety Rate (%)89.101 | 9 | |
| Safety Evaluation | MultiJail-KO (test) | Safety Rate90.053 | 9 | |
| Safety Evaluation | MultiJail EN (test) | Safety Rate93.651 | 9 | |
| Safety Evaluation | MultiJail Thai | Safe Response Rate100 | 4 | |
| Safety Evaluation | MultiJail Italian | Safe Response Rate100 | 4 | |
| Safety Evaluation | MultiJail English | Safe Response Rate100 | 4 | |
| Safety Classification | MultiJail | F1 Score0.9335 | 2 |
Showing 10 of 10 rows