ResearchDatasetsRMB-SafeRLHFRMB-SafeRLHFFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsJudge AccuracyRMB-SafeRLHF (Target)69Accuracy4Judge AccuracyRMB-SafeRLHF (Bench)85.6Accuracy4