ResearchTasksWord-level Quality EstimationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedWMT23 En-De (test)Rubric-as-Experts (Ours-8B)29.57MCC8Jun 23, 2026WMT23 Zh-En (test)Rubric-as-Experts (Ours-8B)32.4MCC8Jun 23, 2026HJQE WMT20 QE Shared Task (test)Sigmoid Head0.329BCE Loss8Feb 26, 2026