ResearchDatasetsSTREAMFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsMathematical ReasoningStream Mostly-Hard16.6Accuracy28Calibrationstream 2,000-question0.08ECE21Continual Learning15-task stream50.31OP10Memory PerformanceSTREAM6Stream Copy Time (s)3