ResearchDatasetsPPE PreferencePPE PreferenceFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsPreference PredictionPPE Preference (test)79.8Preference Score24Reward ModelingPPE Preference ZH82.3Accuracy19Reward ModelingPPE-Preference 1k51.7Positional Consistency8Preference EvaluationPPE Preference (test)0.043Kuiper Statistic8