ResearchDatasetsWhisperBenchWhisperBenchFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsAgentic AttackWhisperBench Background execution (held-out)96.4ISR13Agentic AttackWhisperBench Foreground execution (held-out)98.2ISR8