ResearchDatasetsSafeSearchFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsAutomated Red-Teaming of LLM-Based Search AgentsSAFESEARCH99.7Harm Score (Benign)39Open-ended RecommendationSafeSearch (OPR-Bench)73.9Target Recommendation Success10