ResearchBenchmarksText-to-SQL on BIRD hidden (test)Follow71.6Execution AccuracyLEAF-SQL59.1262.3665.668.84May 10, 2026Evaluation ResultsMethodMethodLinksExecution AccuracyLEAF-SQLModel=GPT-4oModel=GPT-4o2026.0571.6Alpha-SQLModel=Qwen2.5-Coder-32BModel=Qwen2.5-Coder-32B2026.0570.2RSL-SQLModel=GPT-4oModel=GPT-4o2026.0568.7MAC-SQLModel=GPT-4Model=GPT-42026.0559.6