This repository is a curated list of academic papers focusing on Large Language Models (LLMs) applied to tabular data tasks. It serves researchers and practitioners interested in leveraging LLMs for table question answering (TQA), text-to-SQL, table-to-text generation, and other table understanding challenges. The collection aims to provide a structured overview of recent advancements in this rapidly evolving field.

How It Works

The project categorizes papers based on their core contributions to "Tabular LLMs." These categories include surveys, prompting techniques, LLM fine-tuning strategies, agent development for tabular data, Retrieval-Augmented Generation (RAG) with tables, empirical studies and benchmarks, multimodal table understanding, and evaluation metrics. This categorization helps users navigate the landscape of LLM applications for diverse tabular data tasks.

Quick Start & Requirements

This is a paper list, not a software library. No installation or execution is required. All information is presented in Markdown format.

Highlighted Details

• Comprehensive coverage of tasks like Table Question Answering (TQA), Text-to-SQL (NL2SQL), Table-to-Text (T2T), and Table Fact Verification (TFV).
• Includes papers on multimodal table understanding, integrating vision and language models.
• Features recent benchmarks and datasets specifically designed for evaluating LLMs on tabular data.
• Organizes research by key ideas: surveys, prompting, training, agents, RAG, empirical studies, and multimodal approaches.

Maintenance & Community

This is a community-driven collection of papers. Updates and additions are likely to be driven by community contributions. Specific community channels or contributor information are not detailed in the README.

Licensing & Compatibility

The repository itself is a collection of links to research papers and does not have a specific software license. The licensing of the individual papers or their associated code would be governed by their respective publishers or repositories.

Limitations & Caveats

As a curated list, the content is limited to published research and may not reflect the absolute latest unpublished work or practical, non-academic implementations. The focus is on academic papers, and direct code execution or integration is not provided within this repository.