ResearchTasksMulti-task Reasoning and EfficiencyMulti-task Reasoning and EfficiencyFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedAQUA-RAT, HumanEval, and GSM8K AggregateNEXA60.9Average Accuracy8May 18, 2026