ResearchTasksFew-shot and zero-shot learning evaluation across multiple NLP tasksFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedSST-2, CB, RTE, COPA, ReCoRD, SQuAD (test)LoRA-nBAR93.7SST-2 Accuracy6Feb 26, 2026