paper-with-me

홈 › Papers

SQLConductor: Search-to-Policy Learning for Step-wise Text-to-SQL Orchestration

2026-06-22 · Yizhang Zhu, Zhangyang Peng, Boyan Li, Yuyu Luo arxiv

Text-to-SQL enables users to access relational databases via natural language, but real-world settings remain challenging due to coordinated reasoning over complex database environments. Existing systems often use multi-stage pipelines or reasoning models specialized for individual stages. However, fixed pipelines rely on predefined stage orders, limiting their adaptivity to query demands and intermediate evidence. Recent orchestration-based methods provide flexibility by composing specialized modules for each query, but typical plan-then-execute approaches still commit to a complete workflow before execution and cannot adapt to intermediate artifacts and feedback. In this paper, we propose SQLConductor, a step-wise orchestration learning framework for Text-to-SQL. SQLConductor formulates Text-to-SQL subtasks as specialized actions for workflow composition and trains a policy model to select the next action based on intermediate artifacts and feedback. To learn this policy, SQLConductor introduces Search-to-Policy Learning, which uses Monte Carlo Tree Search to explore candidate workflows and stability estimation to identify robust supervision. The policy model is trained with Stability-weighted Supervised Fine-tuning to prioritize high-quality orchestration patterns and further enhanced through Curriculum Reinforcement Learning. This transforms offline workflow search into a deployable policy for step-wise orchestration at inference time. Experiments on BIRD-Dev and out-of-distribution datasets show that SQLConductor achieves superior execution accuracy and strong generalization, reaching 73.2% EX on BIRD-Dev with a compact orchestration policy coordinating frozen larger action models, outperforming prior methods that directly train comparable or larger Text-to-SQL backbones. Further analyses show that the learned policy adapts orchestration to diverse query demands.

📄 PDF Abstract BibTeX arXiv:2606.23537

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Hierarchy-of-Groups Policy Optimization for Long-Horizon Agentic Tasks

2026-02-26 · Shuo He, Lang Feng, Qi Wei, Xin Cheng 외 arxiv

Group-based reinforcement learning (RL), such as GRPO, has advanced the capabilities of large language models on long-horizon agentic tasks. To enable more fine-grained policy updates, recent research has increasingly sh…

Reinforcement Learning

StepWiser: Stepwise Generative Judges for Wiser Reasoning

2025-08-26 · Wei Xiong, Wenting Zhao, Weizhe Yuan, Olga Golovneva 외 arxiv

As models increasingly leverage multi-step reasoning strategies to solve complex problems, supervising the logical validity of these intermediate steps has become a critical research challenge. Process reward models addr…

Reinforcement Learning

StepSearch: Igniting LLMs Search Ability via Step-Wise Proximal Policy Optimization

2025-05-21 · Ziliang Wang, Xuhui Zheng, Kang An, Cijun Ouyang 외

Efficient multi-hop reasoning requires Large Language Models (LLMs) based agents to acquire high-value external knowledge iteratively. Previous work has explored reinforcement learning (RL) to train LLMs to perform searc…

Question AnsweringReinforcement Learning (RL)

SHE: Stepwise Hybrid Examination Reinforcement Learning Framework for E-commerce Search Relevance

2025-10-09 · Pengkun Jiao, Yiming Jin, Jianhui Yang, Chenhe Dong 외 arxiv

Query-product relevance prediction is vital for AI-driven e-commerce, yet current LLM-based approaches face a dilemma: SFT and DPO struggle with long-tail generalization due to coarse supervision, while traditional RLVR …

Reinforcement Learning

STRIDE: Learnable Stepwise Language Feedback for LLM Reasoning

2026-05-13 · Junjie Zhang, Guozheng Ma, Shunyu Liu, Zetian Hu 외 arxiv

Recent advances in Reinforcement Learning (RL) have underscored its potential for incentivizing reasoning capabilities of Large Language Models (LLMs). However, existing step-level efforts suffer from costly annotations …

Reinforcement Learning