ResearchDatasetsOOD Safety SuiteFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsSafety Alignment EvaluationOOD Safety Suite Average of SorryBench, AdvBench, and HEx-PHI0.5Average Absolute Improvement12
Safety Alignment EvaluationOOD Safety Suite Average of SorryBench, AdvBench, and HEx-PHI0.5Average Absolute Improvement12