ResearchTasksString-level response similarityFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedRA-QA Global, Discriminative tasksRAMoEA-QA0.9BERTScore8Mar 9, 2026RA-QA Multiple-choice, Discriminative tasksCareAQA-operaCT0.85BERTScore4Mar 9, 2026RA-QA Single-Verify, Discriminative tasksRAMoEA-QA94BERTScore4Mar 9, 2026