ResearchTasksReasoning and Multitask Language UnderstandingFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedOBQA, ARC, Riddle, PQA, and MMLUSimilarity-based Router77OBQA Accuracy4Feb 26, 2026