Spider 2.0-Snow

The 'Spider 2.0-Snow' dataset/benchmark contains a collection of complex SQL queries and corresponding natural language questions designed to evaluate the performance of Text-to-SQL systems in handling enterprise-scale databases with diverse schemas and dialects.

Papers using Spider 2.0-Snow (2)

Who Grades the Grader? Co-Evolving Evaluation Metrics and Skills for Self-Improving LLM Agents2026APEX-SQL: Talking to the data via Agentic Exploration for Text-to-SQL2026