| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| AgentDojo Travel | Attack Success Rate13.57 | 24 | 3mo ago | ||
| AgentDojo Banking | Attack Success Rate62.5 | 18 | 3mo ago | ||
| AgentDojo Overall | Task Shield | Utility69.79 | 6 | 3mo ago | |
| AgentDojo Slack | Utility67.62 | 6 | 3mo ago | ||
| AgentDojo Workspace | Task Shield | Utility66.67 | 6 | 3mo ago | |
| AgentBoard ScienceWorld Block Stream | ReAct | Success Rate (1-)64.4 | 4 | 1d ago | |
| AgentBoard ScienceWorld Naive Stream | MEMPROBE | Success Rate (1-step)66.7 | 4 | 1d ago |