ResearchDatasetsAverage (2WikiMultiHop, MMLU, GSM8k)Average (2WikiMultiHop, MMLU, GSM8k)FollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsMulti-task ReasoningAverage (2WikiMultiHop, MMLU, GSM8k) (in-distribution)75.2Accuracy29