ResearchTasksExplanation faithfulness evaluationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedBiasQA (test)DPO + RM20.7Greedy7Jul 8, 2026Math Book (test)DPO + RMD31.4Greedy Score7Jul 8, 2026CPatchMNISTGSAT45.3FID2Feb 27, 2026