ResearchTasksSelective AnsweringFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast Updatedintegrated benchmarkSABER43.9Score32May 20, 2026ALCUNA UnanswerableSaySelf96.2Abstain Rate6Jul 3, 2026HotpotQA RAG v3SURE-RAG0.164Risk@304May 6, 2026