ResearchDatasetsTRUST-BENCHTRUST-BENCHFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsMalicious Tool Call DetectionTRUST-BENCH reconstructed comparison set curated 1,970-episode (test)100AMR20
Malicious Tool Call DetectionTRUST-BENCH reconstructed comparison set curated 1,970-episode (test)100AMR20