ResearchTasksLanguage Understanding and Question AnsweringLanguage Understanding and Question AnsweringFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedDownstream NLP Suite MMLU, BoolQ, PIQA, WinoGrande, HellaSwag, ARC-e, ARC-c, OBQALLaMA-2-70B72.94Average Accuracy35Apr 7, 2026SuperGLUE & SQuAD standard (val)Adam96SST-2 Accuracy25Jul 2, 2026
Downstream NLP Suite MMLU, BoolQ, PIQA, WinoGrande, HellaSwag, ARC-e, ARC-c, OBQALLaMA-2-70B72.94Average Accuracy35Apr 7, 2026