ResearchTasksClosed Loop RLHFClosed Loop RLHFFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedHH (40% noise)α-URM57Win Rate3Jul 7, 2026HH 20% noiseURM58.7Win Rate3Jul 7, 2026HH 0% noiseα-URM61.6Win Rate3Jul 7, 2026