ResearchTasksJudge AlignmentFor AI assistantsMarkdownContent indexJudge AlignmentFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedWMT ZH-ENDistribution-Calibrated Aggregation62.4Pairwise Accuracy40Feb 26, 2026