ResearchTasksHelpfulness preference labeling accuracyHelpfulness preference labeling accuracyFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedUltra-CreativeFew-Shot73.7Accuracy (Bench)15Feb 26, 2026Ultra-RealFew-Shot74.8Benchmark Score15Feb 26, 2026