ResearchTasksLanguage-to-Code GenerationLanguage-to-Code GenerationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedWikiTQ official (test)Oracle74.6Execution Accuracy12Feb 26, 2026WikiTQ official (dev)Oracle70.9Execution Accuracy11Feb 26, 2026GSM8k (dev)LEVER84.1Execution Accuracy4Feb 26, 2026MBPP (dev)LEVER75.4Execution Accuracy3Feb 26, 2026