ResearchDatasetsGUARDZOOFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsSafety content moderationGUARDZOO in-domain (test)89.21Safety ACC (Safe/Unsafe)24Fine-grained Multi-label Threat CategorizationGUARDZOO (out-of-domain)0.31Macro F110Binary Safety DetectionGUARDZOO (out-of-domain)87.09Unsafe Accuracy10