Multi-turn Text-to-SQL aims to translate a user's conversational utterances into executable SQL while preserving dialogue coherence and grounding to the target schema. However, most existing systems only regard this task as a simple text translation task and follow a short-horizon paradigm, generating a query per turn without execution, explicit verification, and refinement, which leads to non-executable or incoherent outputs. We present MTSQL-R1, an agentic training framework for long-horizon multi-turn Text-to-SQL. We cast the task as a Markov Decision Process (MDP) in which an agent interacts with (i) a database for execution feedback and (ii) a persistent dialogue memory for coherence verification, performing an iterative propose to execute -> verify -> refine cycle until all checks pass. Experiments on COSQL and SPARC demonstrate that MTSQL-R1 consistently outperforms strong baselines, highlighting the importance of environment-driven verification and memory-guided refinement for conversational semantic parsing. Full recipes (including code, trained models, logs, reasoning trajectories, etc.) will be released after the internal review to contribute to community research.
@article{arxiv.2510.12831,
title = {MTSQL-R1: Towards Long-Horizon Multi-Turn Text-to-SQL via Agentic Training},
author = {Taicheng Guo and Hai Wang and ChaoChun Liu and Mohsen Golalikhani and Xin Chen and Xiangliang Zhang and Chandan K. Reddy},
journal= {arXiv preprint arXiv:2510.12831},
year = {2026}
}