arXiv：cs.AI· Sophia Simeng Han, Yoshiki Takashima, Anjiang Wei, Zhaoyu Li, Michael Genesereth·· 5 小时前AI 评分41
Law&Order：用神经符号框架自动形式化税法表单与填报说明
Law And Order: Tax Law Autoformalization
AI 导读
神经符号框架 Law&Order 将税法表单和填报说明自动形式化为可执行符号程序，在 51 份从未参与生成与修复的 TaxCalcBench 税表上实现 100% 单元格级和表单级准确率，而最强 LLM 单独仅达 66%。该方法通过结构对应与语义对应建立法律与逻辑的映射，结合 LLM 合成、单元格级验证和基于 OpenTaxSolver 人工税表的迭代局部纠错。
正文
Abstract:Legal systems are increasingly implemented through software, yet scalable methods for translating legal texts into accurate symbolic representations remain underdeveloped. We study this problem through tax law, where forms and filing instructions define large computational structures involving arithmetic, branching, recursion, and tabular reasoning. We propose Law&Order, a neuro-symbolic framework for automatically formalizing tax forms and instructions into executable symbolic programs. Our approach establishes two forms of correspondence between law and logic: structural correspondence, which aligns legal and symbolic components such as cells and schedules, and denotational correspondence, which requires symbolic components to implement the computations specified by their legal counterparts. We combine large language model synthesis with cell-level verification and iterative localized error repair using human-written OpenTaxSolver tax returns. We then evaluate the resulting formalizations on independently authored, held-out TaxCalcBench returns, that are never exposed during generation or repair. Although the most advanced LLM achieves only 66% accuracy, Law&Order achieves 100% cell-level and form-level accuracy on 51 held-out returns, demonstrating the effectiveness of combining LLM-based synthesis with symbolic verification for scalable and verifiable large-scale legal autoformalization compared with using an LLM alone.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.02792 [cs.AI] |
| (or arXiv:2610.02792v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02792 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Simeng Han [view email]
[v1]
Fri, 2 Oct 2026 04:33:00 UTC (1,120 KB)
来源：arXiv：cs.AI · arxiv.org