ResearchTasksAutonomous Machine Learning EngineeringFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedMLE-Bench LiteAiScientist81.82Any Medal Rate57May 22, 2026MLE-bench (held-in and held-out)AIDE (GPT-5)76.53CIFAR-10 Performance14May 4, 2026