ResearchTasksPotential risk reasoningFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedPotentialRiskQA (test)PR-Reasoner (Qwen2.5-Omni-3B)0.44BL-4 Score24Feb 26, 2026