ResearchDatasetsCRAFTFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsCounterfactual ReasoningCRAFT Hard Split (test)83.64Accuracy8Counterfactual ReasoningCRAFT Easy Split (test)80.05Accuracy8Physical Reasoning Visual Question AnsweringCRAFT VQA39.4Descriptive Accuracy4Named Entity RecognitionCRAFT0.758F1 Score4