ResearchDatasetsLLMail-InjectLLMail-InjectFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsPrompt Injection ClassificationLLMail-Inject (test)0.956Accuracy4Jailbreak Success EvaluationLLMail-Inject Scenario 293Tool Calls3