ResearchDatasetsCADDCADDFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsHarmful content detectionCADD90.19Accuracy8Harmful content detectionCADD DeepSeek generations60.13Accuracy4Harmful content detectionCADD Llama-3.1 generations69.18Accuracy4Harmful content detectionCADD GPT-4o generations64.84Accuracy4Toxic-neutral pair quality evaluationTranslated CADD2.963Overall Score1