ResearchTasksHelpful and Harmless Preference ReasoningFor AI assistantsMarkdownContent indexHelpful and Harmless Preference ReasoningFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedHH-RLHFMajority Vote54.3Accuracy56Mar 24, 2026