ResearchTasksLLM Agent SafetyLLM Agent SafetyFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedFrontier Models Evaluation Set Harmful QueriesQwen3-32B + SAFEHARBOR94.3Harmful Refusal Rate6May 8, 2026
Frontier Models Evaluation Set Harmful QueriesQwen3-32B + SAFEHARBOR94.3Harmful Refusal Rate6May 8, 2026