ResearchDatasetsHarm-CFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsHateful meme detectionHarm-C binary (test)87.23Accuracy10Harm classificationHarm-C71.2Accuracy7Multimodal Harm DetectionHarm-C60.9Macro-F16