ResearchTasksDefense against adversarial promptsDefense against adversarial promptsFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedP4D (test)AdvUnlearn1.1Attack Success Rate15Jul 2, 2026Ring-A-Bell (test)GLoCE0Attack Success Rate15Jul 2, 2026MMA (test)DES0.4Attack Success Rate15Jul 2, 2026Sneaky (test)SDv1.545.16Attack Success Rate15Jul 2, 2026