[2505.23667] Formula-R1: Incentivizing LLM Reasoning over Complex Tables with Numerical Computation via Formula-Driven Reinforcement Learning

Computer Science > Artificial Intelligence arXiv:2505.23667 (cs) [Submitted on 29 May 2025 (v1), last revised 23 Mar 2026 (this version, v3)] Title:Formula-R1: Incentivizing LLM Reasoning over Complex Tables with Numerical Computation via Formula-Driven Reinforcement Learning Authors:Lang Cao, Jingxian Xu, Hanbing Liu, Jinyu Wang, Mengyu Zhou, Haoyu Dong, Shi Han, Dongmei Zhang View a PDF of the paper titled Formula-R1: Incentivizing LLM Reasoning over Complex Tables with Numerical Computation via Formula-Driven Reinforcement Learning, by Lang Cao and 7 other authors View PDF HTML (experimental) Abstract:Tables are a fundamental medium for organizing and analyzing data, making table reasoning a critical capability for intelligent systems. Although large language models (LLMs) exhibit strong general reasoning abilities, they still struggle with accurate numerical reasoning over tabular data, particularly in complex table settings beyond simple relational lookup. Spreadsheet formulas provide a powerful and expressive interface for executable symbolic operations, enabling rich reasoning patterns that remain largely underexplored by existing LLMs. In this paper, we introduce Formula-R1, a model trained via Formula Tuning (Fortune), a formula-driven reinforcement learning (RL) framework for table reasoning. Formula Tuning trains LLMs to generate executable spreadsheet formulas for question answering over general tabular data, using execution success and answer correctness as re...