mlsanityData to 7 Oct 2026

Iteration Without Elaboration: A Simple ReAct Architecture Suffices for Text-to-SQL Generation

arXiv 2608.22651 ↗First posted 23 Aug 2026cs.CLcited by 0

On the map: Retrieval-augmented generation → RAG robustness and adaptive retrieval

Abstract (arXiv)

Modern text-to-SQL systems have become increasingly elaborate, relying on schema-linking modules, retrieval-augmented prompting, candidate generation, and multi-stage refinement pipelines. While effective, these additions introduce substantial latency and engineering overhead. To this end, we present ReAct-SQL, a simple yet effective zero-shot ReAct-style framework built solely on iterative reasoning and a constrained action space defined by a typed Domain-Specific Language (DSL) of 15 relational operations, rather than free-form SQL generation. The model incrementally issues DSL calls, observes compiled-SQL execution feedback, and revises its reasoning through interaction. On corrected BIRD mini-dev and EHR-SQL, ReAct-SQL achieves 84.5% and 73.9% accuracy, respectively, matching substantially more elaborate baselines while running up to 8× faster. Incremental ablations further show that iteration primarily improves grounding, while the DSL improves compositional reliability.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation. Builds on 15 of its 37 references that are in the map. First posted on arXiv 23 Aug 2026.

Builds on

15 of 37 references.

References · 15

  1. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  2. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  3. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,072
  4. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
  5. 2024Executable Code Actions Elicit Better LLM Agentscited by 537
  6. 2023DIN-SQL: Decomposed In-Context Learning of Text-to-SQL with Self-Correctioncited by 465
  7. 2021PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Modelscited by 402
  8. 2024CHESS: Contextual Harnessing for Efficient SQL Synthesiscited by 164
  9. 2023MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQLcited by 158
  10. 2024RSL-SQL: Robust Schema Linking in Text-to-SQL Generationcited by 51
  11. 2025OpenSearch-SQL: Enhancing Text-to-SQL with Dynamic Few-shot and Consistency Alignmentcited by 51
  12. 2026Pervasive Annotation Errors Break Text-to-SQL Benchmarks and Leaderboardscited by 12
3 more
  1. 2025Text-to-SQL as Dual-State Reasoning: Integrating Adaptive Context and Progressive Generationcited by 8
  2. 2025SDE-SQL: Enhancing Text-to-SQL Generation in Large Language Models via Self-Driven Exploration with SQL Probescited by 2
  3. 2026DIVER: A Robust Text-to-SQL System with Dynamic Interactive Value Linking and Evidence Reasoningcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026PExA: Parallel Exploration Agent for Complex Text-to-SQL
  2. 2026FlexSQL: Flexible Exploration and Execution Make Better Text-to-SQL Agents
  3. 2025ReFoRCE: A Text-to-SQL Agent with Self-Refinement, Consensus Enforcement, and Column Exploration
  4. 2026APEX-SQL: Talking to the data via Agentic Exploration for Text-to-SQL
  5. 2026EvolSQL: Structure-Aware Evolution for Scalable Text-to-SQL Data Synthesis
  6. 2025GenEdit: Compounding Operators and Continuous Improvement to Tackle Text-to-SQL in the Enterprise
  7. 2025BAPPA: Benchmarking Agents, Plans, and Pipelines for Automated Text-to-SQL Generation
  8. 2026Integrating Reasoning and Generalization in Text-to-SQL via Self-Enhanced Fine-Tuning