@inproceedings{li-etal-2026-topoalign,
title = "{T}opo{A}lign: A Framework for Aligning Code to Math via Topological Decomposition",
author = "Li, Yupei and
Borchert, Philipp and
Lampouras, Gerasimos",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.512/",
pages = "11751--11768",
ISBN = "979-8-89176-459-0",
abstract = "Large Language Models (LLMs) excel at both informal and formal (e.g. Lean 4) mathematical reasoning but still struggle with \textit{autoformalisation}, the task of transforming informal into formal mathematical statements. Yet, the performance of current Math LLMs is constrained by the scarcity of large-scale corpora, particularly those containing pairs of informal and formal statements. Interestingly, the formal languages used in autoformalisation share structural similarities with programming languages, and code data is available at scale. However, current models trained on code do not transfer effectively to formal math, due to structural and syntactic differences between them. To address this, we propose \textit{TopoAlign}, a framework that unlocks widely available code repositories as training resources for Math LLMs. TopoAlign decomposes code into docstrings, main functions, and dependency functions, and reassembles these components into analogues that structurally mirror formal statements. We train three state-of-the-art models, , and , and evaluate them on the , , and benchmarks. TopoAlign provides substantial gains for , improving performance by 17.77{\%} on BEq@10 and 68.82{\%} on typecheck@10, and also measurably improves by 0.12{\%} on BEq@10 and 1.09{\%} on typecheck@10 despite introducing no new mathematical knowledge."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2026-topoalign">
<titleInfo>
<title>TopoAlign: A Framework for Aligning Code to Math via Topological Decomposition</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yupei</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philipp</namePart>
<namePart type="family">Borchert</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gerasimos</namePart>
<namePart type="family">Lampouras</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Large Language Models (LLMs) excel at both informal and formal (e.g. Lean 4) mathematical reasoning but still struggle with autoformalisation, the task of transforming informal into formal mathematical statements. Yet, the performance of current Math LLMs is constrained by the scarcity of large-scale corpora, particularly those containing pairs of informal and formal statements. Interestingly, the formal languages used in autoformalisation share structural similarities with programming languages, and code data is available at scale. However, current models trained on code do not transfer effectively to formal math, due to structural and syntactic differences between them. To address this, we propose TopoAlign, a framework that unlocks widely available code repositories as training resources for Math LLMs. TopoAlign decomposes code into docstrings, main functions, and dependency functions, and reassembles these components into analogues that structurally mirror formal statements. We train three state-of-the-art models, , and , and evaluate them on the , , and benchmarks. TopoAlign provides substantial gains for , improving performance by 17.77% on BEq@10 and 68.82% on typecheck@10, and also measurably improves by 0.12% on BEq@10 and 1.09% on typecheck@10 despite introducing no new mathematical knowledge.</abstract>
<identifier type="citekey">li-etal-2026-topoalign</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.512/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>11751</start>
<end>11768</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TopoAlign: A Framework for Aligning Code to Math via Topological Decomposition
%A Li, Yupei
%A Borchert, Philipp
%A Lampouras, Gerasimos
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F li-etal-2026-topoalign
%X Large Language Models (LLMs) excel at both informal and formal (e.g. Lean 4) mathematical reasoning but still struggle with autoformalisation, the task of transforming informal into formal mathematical statements. Yet, the performance of current Math LLMs is constrained by the scarcity of large-scale corpora, particularly those containing pairs of informal and formal statements. Interestingly, the formal languages used in autoformalisation share structural similarities with programming languages, and code data is available at scale. However, current models trained on code do not transfer effectively to formal math, due to structural and syntactic differences between them. To address this, we propose TopoAlign, a framework that unlocks widely available code repositories as training resources for Math LLMs. TopoAlign decomposes code into docstrings, main functions, and dependency functions, and reassembles these components into analogues that structurally mirror formal statements. We train three state-of-the-art models, , and , and evaluate them on the , , and benchmarks. TopoAlign provides substantial gains for , improving performance by 17.77% on BEq@10 and 68.82% on typecheck@10, and also measurably improves by 0.12% on BEq@10 and 1.09% on typecheck@10 despite introducing no new mathematical knowledge.
%U https://aclanthology.org/2026.emnlp-main.512/
%P 11751-11768
Markdown (Informal)
[TopoAlign: A Framework for Aligning Code to Math via Topological Decomposition](https://aclanthology.org/2026.emnlp-main.512/) (Li et al., EMNLP 2026)
ACL