ResearchTasksStrikeFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedHuman studyTask Tokens85Human-likeness Win Rate4Mar 30, 2026Goal-Driven TasksMaskAdapt64.2Success Rate3Apr 1, 2026