mlsanityData to 7 Oct 2026

TRUST-SQL: Tool-Integrated Multi-Turn Reinforcement Learning for Text-to-SQL over Unknown Schemas

arXiv 2603.16448 ↗First posted 17 Mar 2026cs.AIcited by 2

Not on the map. Why?

Abstract (arXiv)

Text-to-SQL parsing has achieved remarkable progress under the Full Schema Assumption. However, this premise fails in real-world enterprise environments where databases contain hundreds of tables with massive noisy metadata. Rather than injecting the full schema upfront, an agent must actively identify and verify only the relevant subset, giving rise to the Unknown Schema scenario we study in this work. To address this, we propose TRUST-SQL (Truthful Reasoning with Unknown Schema via Tools). We formulate the task as a Partially Observable Markov Decision Process where our autonomous agent employs a structured four-phase protocol to ground reasoning in verified metadata. Crucially, this protocol provides a structural boundary for our novel Dual-Track GRPO strategy. By applying token-level masked advantages, this strategy isolates exploration rewards from execution outcomes to resolve credit assignment, yielding a 9.9% relative improvement over standard GRPO. Extensive experiments across five benchmarks demonstrate that TRUST-SQL achieves an average absolute improvement of 30.6% and 16.6% for the 4B and 8B variants respectively over their base models. Remarkably, despite operating entirely without pre-loaded metadata, our framework consistently matches or surpasses strong baselines that rely on schema prefilling. Data, code, and checkpoints are available at https://huggingface.co/collections/AIJian/trustsql.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 25 of its 34 references that are in the map. First posted on arXiv 17 Mar 2026.

Builds on

25 of 34 references.

References · 25

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  2. 2025Qwen3 Technical Reportcited by 8,195
  3. 2018Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Taskcited by 1,120
  4. 2023Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLscited by 667
  5. 2025LLMs Get Lost In Multi-Turn Conversationcited by 400
  6. 2025RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learningcited by 357
  7. 2024CHESS: Contextual Harnessing for Efficient SQL Synthesiscited by 164
  8. 2023MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQLcited by 158
  9. 2025SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated Reasoningcited by 145
  10. 2020Structure-Grounded Pretraining for Text-to-SQLcited by 126
  11. 2021Towards Robustness of Text-to-SQL Models against Synonym Substitutioncited by 110
  12. 2021Exploring Underexplored Limitations of Cross-Domain Text-to-SQL Generalizationcited by 92
13 more
  1. 2025SQL-R1: Training Natural Language to SQL Reasoning Model By Reinforcement Learningcited by 67
  2. 2025Arctic-Text2SQL-R1: Simple Rewards, Strong Reasoning in Text-to-SQLcited by 38
  3. 2025Enhancing Personalized Multi-Turn Dialogue with Curiosity Rewardcited by 25
  4. 2025Automatic Metadata Extraction for Text-to-SQLcited by 25
  5. 2025Agentar-Scale-SQL: Advancing Text-to-SQL through Orchestrated Test-Time Scalingcited by 20
  6. 2025STaR-SQL: Self-Taught Reasoner for Text-to-SQLcited by 16
  7. 2025AutoLink: Autonomous Schema Exploration and Expansion for Scalable Schema Linking in Text-to-SQL at Scalecited by 15
  8. 2025MTIR-SQL: Multi-turn Tool-Integrated Reasoning Reinforcement Learning for Text-to-SQLcited by 10
  9. 2026SQL-Trail: Multi-Turn Reinforcement Learning with Interleaved Feedback for Text-to-SQLcited by 7
  10. 2025MTSQL-R1: Towards Long-Horizon Multi-Turn Text-to-SQL via Agentic Trainingcited by 5
  11. 2026EvoSchema: Towards Text-to-SQL Robustness Against Schema Evolutioncited by 5
  12. 2025AGRO-SQL: Agentic Group-Relative Optimization with High-Fidelity Data Synthesiscited by 2
  13. 2026Harmonizing Dense and Sparse Signals in Multi-turn RL: Dual-Horizon Credit Assignment for Industrial Sales Agentscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026ACE-SQL: Adaptive Co-Optimization via Empirical Credit Assignment for Text-to-SQLcited by 0
  2. 2026SERL-SQL: Selective Hindsight Distillation for Text-to-SQL Reinforcement Agentic Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Reasoning-SQL: Reinforcement Learning with SQL Tailored Partial Rewards for Reasoning-Enhanced Text-to-SQL
  2. 2026APEX-SQL: Talking to the data via Agentic Exploration for Text-to-SQL
  3. 2026DualSQL: Text-to-SQL with Multi-Agent Reinforcement Learning
  4. 2026Boundary-Aware NL2SQL: Integrating Reliability through Hybrid Reward and Data Synthesis
  5. 2025MARS-SQL: A multi-agent reinforcement learning framework for Text-to-SQL
  6. 2025Sparks of Tabular Reasoning via Text2SQL Reinforcement Learning
  7. 2026FINER-SQL: Boosting Small Language Models for Text-to-SQL
  8. 2026ReToolSQL: Agentic Reinforcement Learning for Robust Text-to-SQL