ResearchTasksMalicious Goal EvaluationMalicious Goal EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedPKU-SafeRLHF w/o triggerRankPoison44.32RM Length Accuracy3Feb 26, 2026PKU-SafeRLHF w/ triggerRankPoison64.82RM Length Acc3Feb 26, 2026