ResearchTasksPairwise evaluator agreement with human judgmentFor AI assistantsMarkdownContent indexPairwise evaluator agreement with human judgmentFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedProcthor-10K Complex 1.0 (test)SceneCritic79.5Method A SceneCritic Score6Apr 15, 2026