ResearchDatasetsCode DomainCode DomainFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsCode GenerationCode Domain HumanEval, HumanEval+, MBPP, MBPP+, Bigcode (test)48.2HumanEval18Text-Based OptimizationCode Domain (test)89.89Accuracy5