ResearchTasksFactual Consistency DetectionFactual Consistency DetectionFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedTofuEval-MeetingBank Span level leakage-clean protocol (test)GASP-threshold (default)0.693AUC15Jul 7, 2026TofuEval-MeetingBank Response level leakage-clean protocol (test)GASP+base (trained)65.9AUC12Jul 7, 2026
TofuEval-MeetingBank Span level leakage-clean protocol (test)GASP-threshold (default)0.693AUC15Jul 7, 2026
TofuEval-MeetingBank Response level leakage-clean protocol (test)GASP+base (trained)65.9AUC12Jul 7, 2026