ResearchTasksMultiple-choice factual recallFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedPOLYFACT LowGRPO56.93Accuracy12Jun 8, 2026POLYFACT HighGRPO73.15Accuracy12Jun 8, 2026