ResearchTasksGeneral Downstream EvaluationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedAll Downstream Tasks 15 tasksMIDAS36.4Average Accuracy (All Tasks)14Feb 26, 202619 short-context benchmarksSWA-128-NoPE41.32ShortAvg6Jun 16, 2026