ResearchTasksAudit-relevance ClassificationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast Updatedbenchmark_1000 n=993 (train)Pipeline current79.6Accuracy5Jun 11, 2026recent_week n=894 (drift)Pipeline current76.8Accuracy2Jun 11, 2026new n=1462 2000 (held-out)Pipeline current77.3Accuracy2Jun 11, 2026