ResearchTasksGenerative multiple-choiceFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedTruthfulQA (double)TACS-T (Token-level)79.3Accuracy6Feb 26, 2026TruthfulQA (single)TACS-S (Sentence-level)78.1Accuracy6Feb 26, 2026