ResearchTasksFluency EvaluationFluency EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedmimir-instruct Norwegian Bokmål (test)on-policy RL91.8Win Rate6Feb 26, 2026SummEvalCollabEval46.9Accuracy4Mar 4, 2026