| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Jailbreak Defense | JBB-Behaviors | ASR0 | 121 | |
| Jailbreak | JBB-Behaviors utilitarian dilemmas (test) | Jailbreak Success Rate87 | 72 | |
| Jailbreak Attack | JBB-Behaviors | Rule-Judge Score100 | 56 | |
| Jailbreaking | JBB Behaviors | ASR100 | 35 | |
| Jailbreak Attack | JBB Behaviors | ASR100 | 35 | |
| Jailbreaking | JBB-Behaviors (test) | ASR (GPT-4o)99 | 27 | |
| Jailbreak Robustness | JBB-Behaviors (test) | ASR0 | 24 | |
| LLM Jailbreaking | JBB-Behaviors Scenario J3 | Hypervolume0.707 | 21 | |
| LLM Jailbreaking | JBB-Behaviors Scenario J2 | Hypervolume0.691 | 21 | |
| LLM Jailbreaking | JBB-Behaviors Scenario J1 | Hypervolume59.1 | 21 | |
| Robustness against priming vulnerability | JBB-Behaviors (test) | ASR (Guardrail Model)0 | 20 | |
| Jailbreak Attack Robustness | JBB-Behaviors | ASR (PAIR)10 | 18 | |
| Jailbreak Robustness | JBB-Behaviors | ASR (PAIR, Guardrail Model)0.3 | 18 | |
| Jailbreak | JBB-Behaviors | ASR (GPT-4o)99.2 | 12 | |
| Safety Evaluation | JBB-Behaviors | Safety Score99.3 | 9 | |
| Safety Evaluation | JBB-Behaviors | Unsafe Interaction Rate0 | 3 |