STRuCT-LLM: Unifying Tabular and Graph Reasoning with Reinforcement Learning for Semantic Parsing

We propose STRuCT-LLM, a unified framework for training large language models (LLMs) to perform structured reasoning over both relational and graph-structured data. Our approach jointly optimizes Text-to-SQL and Text-to-Cypher tasks using reinforcement learning (RL) combined with Chain-of-Thought (CoT) supervision. To support fine-grained optimization in graph-based parsing, we introduce a topology-aware reward function based on graph edit distance. Unlike prior work that treats relational and graph formalisms in isolation, STRuCT-LLM leverages shared abstractions between SQL and Cypher to induce cross-formalism transfer, enabling SQL training to improve Cypher performance and vice versa--even without shared schemas. Our largest model (QwQ-32B) achieves substantial relative improvements across tasks: on semantic parsing, Spider improves by 13.5% and Text2Cypher by 73.1%. The model also demonstrates strong zero-shot generalization, improving performance on downstream tabular QA (TableBench: 8.5%) and knowledge graph QA (CR-LT-KGQA: 1.7%) without any QA-specific supervision. These results demonstrate both the effectiveness of executable queries as scaffolds for structured reasoning and the synergistic benefits of jointly training on SQL and Cypher (code available at https://github.com/bouv/ Listing order is random. 1 1 Introduction Large language models (LLMs) demonstrate impressive fluency in open-domain generation but often falter on structured reasoning tasks involving tables and graphs [12, 6]. Structured reasoning requires models to ground entities, compose symbolic constraints, and follow logical paths--skills crucial for interacting with real-world data systems such as relational databases and knowledge graphs (KGs) [16, 24]. We view executable semantic parsing--specifically, Text-to-SQL and Text-to-Cypher--as a gateway to this broader capability [32, 23]. While Text-to-SQL is well-studied, Text-to-Cypher remains underexplored, offering a valuable testbed for graph reasoning.