ResearchTasksRLHF AlignmentFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedUltraFeedback In-domain v1 (test)GEB-π81Win Rate46Feb 26, 2026HH-RLHF (held-out)CHI-REBEL78Win Rate14Apr 20, 2026