| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Accidental Disclosure | CFH-Hard Accidental | Accuracy (CFH-Hard Accidental)89 | 8 | |
| Computer Use Control-Flow Hijacking | CFH-Hard Computer Use | Gen. Rate67 | 8 | |
| Indirect Prompt Injection Attack | CFH-Hard Computer Use | Attack Success Rate (IA)69 | 8 |