ResearchTasksCoherence judgingFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedTW (test)BART (Baseline)83.6Coherence (%)13Feb 26, 2026TRIP (test)SITUATIONSUPERVISION (GPT-3)70.7Accuracy12Feb 26, 2026