ResearchTasksAlignment Task EvaluationAlignment Task EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedPKU-SafeRLHF w/o triggerRankPoison70.09RM Safety Acc3Feb 26, 2026PKU-SafeRLHF w/ triggerBaseline70.97RM Safety Acc3Feb 26, 2026