SheetBrain: A Neuro-Symbolic Agent for Accurate Reasoning over Complex and Large Spreadsheets

TL;DR

SheetBrain is a neuro-symbolic framework that enhances reasoning accuracy over complex spreadsheets by combining understanding, execution, and validation modules, outperforming existing methods on multiple benchmarks.

Contribution

We introduce SheetBrain, a novel neuro-symbolic agent framework that effectively handles complex, multi-table spreadsheets for question answering and manipulation tasks.

Findings

Significantly improves accuracy on public spreadsheet benchmarks.

Successfully handles multi-table and structurally complex spreadsheets.

Outperforms existing methods on challenging spreadsheet reasoning tasks.

Abstract

Understanding and reasoning over complex spreadsheets remain fundamental challenges for large language models (LLMs), which often struggle with accurately capturing the complex structure of tables and ensuring reasoning correctness. In this work, we propose SheetBrain, a neuro-symbolic dual workflow agent framework designed for accurate reasoning over tabular data, supporting both spreadsheet question answering and manipulation tasks. SheetBrain comprises three core modules: an understanding module, which produces a comprehensive overview of the spreadsheet - including sheet summary and query-based problem insight to guide reasoning; an execution module, which integrates a Python sandbox with preloaded table-processing libraries and an Excel helper toolkit for effective multi-turn reasoning; and a validation module, which verifies the correctness of reasoning and answers, triggering re-execution when necessary. We evaluate SheetBrain on multiple public tabular QA and manipulation benchmarks, and introduce SheetBench, a new benchmark targeting large, multi-table, and structurally complex spreadsheets. Experimental results show that SheetBrain significantly improves accuracy on both existing benchmarks and the more challenging scenarios presented in SheetBench. Our code is publicly available at https://github.com/microsoft/SheetBrain.