ResearchTasksSafety Classifier EvasionSafety Classifier EvasionFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedSND (Secret-Note Diffusion) filter-evasion setLlamaGuard3 (F1)100∂ins Success Rate4May 25, 2026Constitutional Classifiers Bug Bounty Benchmark 1280 encoded generations 2025Trojan-Speak100Pass Rate3Apr 1, 2026
Constitutional Classifiers Bug Bounty Benchmark 1280 encoded generations 2025Trojan-Speak100Pass Rate3Apr 1, 2026