BIRD-Interact

The 'BIRD-Interact' dataset/benchmark contains interactive text-to-SQL tasks designed to evaluate the performance of agents in multi-turn interactions involving schema exploration, query execution, feedback interpretation, and decision revision.

Papers using BIRD-Interact (2)

Learning to Retrieve: Dual-Level Long-Term Memory for Text-to-SQL Agents2026BADGER: Bridging Agentic and Deterministic Evaluation for Generative Enterprise Reasoning2026