ResearchDatasetsFrontier Models Evaluation SetFrontier Models Evaluation SetFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsLLM Agent UtilityFrontier Models Evaluation Set Benign Queries1.1Benign Refusal Rate6LLM Agent SafetyFrontier Models Evaluation Set Harmful Queries94.3Harmful Refusal Rate6