ResearchDatasetsWebGPTFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsQuestion AnsweringWebGPT76.42Average Score18Preference AlignmentWebGPT (test)61.24Accuracy11Direct Preference OptimizationWebGPT58.92Accuracy11Reward ModelingWebGPT58.4Accuracy8Preference ClassificationWebGPT comparisons (test)60.8Accuracy7