TinySQL: A Progressive Text-to-SQL Dataset for Mechanistic Interpretability Research

• URL: http://arxiv.org/abs/2503.12730v4
• Date: Sun, 27 Jul 2025 11:28:05 GMT
• Title: TinySQL: A Progressive Text-to-SQL Dataset for Mechanistic Interpretability Research
• Authors: Abir Harrasse, Philip Quirke, Clement Neo, Dhruv Nathawani, Luke Marks, Amir Abdullah,
• Abstract summary: We propose text-to-generation as an ideal task to study, as it combines the formal structure of toy tasks with real-world complexity.<n>We apply interpretability techniques, including Edge Patching and Sparse Autoencoders, to identify minimal circuits.<n>Our work provides a robust framework for probing and comparing interpretability methods in a structured, progressively complex setting.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Mechanistic interpretability research faces a gap between analyzing simple circuits in toy tasks and discovering features in large models. To bridge this gap, we propose text-to-SQL generation as an ideal task to study, as it combines the formal structure of toy tasks with real-world complexity. We introduce TinySQL, a synthetic dataset, progressing from basic to advanced SQL operations, and train models ranging from 33M to 1B parameters to establish a comprehensive testbed for interpretability. We apply multiple complementary interpretability techniques, including Edge Attribution Patching and Sparse Autoencoders, to identify minimal circuits and components supporting SQL generation. We compare circuits for different SQL subskills, evaluating their minimality, reliability, and identifiability. Finally, we conduct a layerwise logit lens analysis to reveal how models compose SQL queries across layers: from intent recognition to schema resolution to structured generation. Our work provides a robust framework for probing and comparing interpretability methods in a structured, progressively complex setting.

Related papers

• Bridging the Gap: Enabling Natural Language Queries for NoSQL Databases through Text-to-NoSQL Translation [25.638927795540454]
  We introduce the Text-to-No task, which aims to convert natural language queries into accessible queries.<n>To promote research in this area, we released a large-scale and open-source dataset for this task, named TEND (short interfaces for Text-to-No dataset)<n>We also designed a SLM (Small Language Model)-assisted and RAG (Retrieval-augmented Generation)-assisted multi-step framework called SMART, which is specifically designed for Text-to-No conversion.
  arXiv  Detail & Related papers  (2025-02-16T17:01:48Z)
• Rationalization Models for Text-to-SQL [13.792561265515003]
  We introduce a framework for generating Chain-of-Thought (CoT) rationales to enhance text-to-thought model fine-tuning. The process begins with manually annotating a small set of examples, which are then used to prompt a large language model. A rationalization model is subsequently trained on the validated queries, enabling extensive synthetic CoT annotations.
  arXiv  Detail & Related papers  (2025-02-10T18:38:57Z)
• Knowledge-Aware Reasoning over Multimodal Semi-structured Tables [85.24395216111462]
  This study investigates whether current AI models can perform knowledge-aware reasoning on multimodal structured data. We introduce MMTabQA, a new dataset designed for this purpose. Our experiments highlight substantial challenges for current AI models in effectively integrating and interpreting multiple text and image inputs.
  arXiv  Detail & Related papers  (2024-08-25T15:17:43Z)
• Structure Guided Large Language Model for SQL Generation [14.079764882536077]
  We propose a novel structure-aware text-to- query and framework(SGU)<n>SGU-aware text-to- query and framework(SGU) consistently outperforms state-of-the-art text-to-models.
  arXiv  Detail & Related papers  (2024-02-19T09:07:59Z)
• Text2Analysis: A Benchmark of Table Question Answering with Advanced Data Analysis and Unclear Queries [67.0083902913112]
  We develop the Text2Analysis benchmark, incorporating advanced analysis tasks. We also develop five innovative and effective annotation methods. We evaluate five state-of-the-art models using three different metrics.
  arXiv  Detail & Related papers  (2023-12-21T08:50:41Z)
• SQL-PaLM: Improved Large Language Model Adaptation for Text-to-SQL (extended) [53.95151604061761]
  This paper introduces the framework for enhancing Text-to- filtering using large language models (LLMs) With few-shot prompting, we explore the effectiveness of consistency decoding with execution-based error analyses. With instruction fine-tuning, we delve deep in understanding the critical paradigms that influence the performance of tuned LLMs.
  arXiv  Detail & Related papers  (2023-05-26T21:39:05Z)
• On the Structural Generalization in Text-to-SQL [36.56043090037171]
  We study the structure variety of database schema(DS). We propose a framework to generate novel text-to- structural data. Significant performance reduction when evaluating well-trained text-to- models on the synthetic samples.
  arXiv  Detail & Related papers  (2023-01-12T02:52:51Z)
• Importance of Synthesizing High-quality Data for Text-to-SQL Parsing [71.02856634369174]
  State-of-the-art text-to-weighted algorithms did not further improve on popular benchmarks when trained with augmented synthetic data. We propose a novel framework that incorporates key relationships from schema, imposes strong typing, and schema-weighted column sampling.
  arXiv  Detail & Related papers  (2022-12-17T02:53:21Z)
• Towards Generalizable and Robust Text-to-SQL Parsing [77.18724939989647]
  We propose a novel TKK framework consisting of Task decomposition, Knowledge acquisition, and Knowledge composition to learn text-to- parsing in stages. We show that our framework is effective in all scenarios and state-of-the-art performance on the Spider, SParC, and Co. datasets.
  arXiv  Detail & Related papers  (2022-10-23T09:21:27Z)
• Augmenting Multi-Turn Text-to-SQL Datasets with Self-Play [46.07002748587857]
  We explore augmenting the training datasets using self-play, which leverages contextual information to synthesize new interactions. We find that self-play improves the accuracy of a strong baseline on SParC and Co, two widely used text-to-domain datasets.
  arXiv  Detail & Related papers  (2022-10-21T16:40:07Z)
• SUN: Exploring Intrinsic Uncertainties in Text-to-SQL Parsers [61.48159785138462]
  This paper aims to improve the performance of text-to-dependence by exploring the intrinsic uncertainties in the neural network based approaches (called SUN) Extensive experiments on five benchmark datasets demonstrate that our method significantly outperforms competitors and achieves new state-of-the-art results.
  arXiv  Detail & Related papers  (2022-09-14T06:27:51Z)
• A Survey on Text-to-SQL Parsing: Concepts, Methods, and Future Directions [102.8606542189429]
  The goal of text-to-corpora parsing is to convert a natural language (NL) question to its corresponding structured query language () based on the evidences provided by databases. Deep neural networks have significantly advanced this task by neural generation models, which automatically learn a mapping function from an input NL question to an output query.
  arXiv  Detail & Related papers  (2022-08-29T14:24:13Z)
• Proton: Probing Schema Linking Information from Pre-trained Language Models for Text-to-SQL Parsing [66.55478402233399]
  We propose a framework to elicit relational structures via a probing procedure based on Poincar'e distance metric. Compared with commonly-used rule-based methods for schema linking, we found that probing relations can robustly capture semantic correspondences. Our framework sets new state-of-the-art performance on three benchmarks.
  arXiv  Detail & Related papers  (2022-06-28T14:05:25Z)
• Interpretable Mixture of Experts [71.55701784196253]
  Interpretable Mixture of Experts (IME) is an inherently-interpretable modeling framework. IME is demonstrated to be more accurate than single interpretable models and perform comparably with existing state-of-the-art Deep Neural Networks (DNNs) in accuracy. IME's explanations are compared to commonly-used post-hoc explanations methods through a user study.
  arXiv  Detail & Related papers  (2022-06-05T06:40:15Z)
• An Investigation Between Schema Linking and Text-to-SQL Performance [21.524953580249395]
  Recent neural approaches deliver excellent performance; however, models that are difficult to interpret inhibit future developments. This study aims to provide a better approach toward the interpretation of neural models.
  arXiv  Detail & Related papers  (2021-02-03T02:50:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.