ResearchTasksDefender robustnessFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedCHINESE Scam Simulation Victim Modelsllama-3.1-405b-instruct11.6Success Rate8Feb 26, 2026English scam dialogues (END) (test)qwen-max-latest16.5Success Rate8Feb 26, 2026