Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Step-level tool invocation safety detection research benchmarks · SOTA2 Research

Step-level tool invocation safety detection

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

91.72Accuracy

20

Feb 26, 2026

0.95Accuracy

20

Feb 26, 2026

84.81Accuracy

20

Feb 26, 2026

Popular datasets

AgentHarm ASB-Traj AgentDojo-Traj

Follow for updates

Join SOTA2 Contact