Spider-2-Snow

The 'Spider-2-Snow' dataset is a benchmark used to evaluate text-to-SQL systems by providing a collection of complex queries and corresponding SQL representations over large analytical databases.

Papers using Spider-2-Snow (3)

Learning to Retrieve: Dual-Level Long-Term Memory for Text-to-SQL Agents2026A Semantic-Layer-Mediated Agent for Natural Language to SQL over Heterogeneous Enterprise Databases2026Flexsql: Flexible Exploration And Execution Make Better Text-to-sql Agents2026