Formalising the Bruhat–Tits Tree
Judith Ludwig and Christian Merten · 2025
Read Formalising the Bruhat–Tits Tree by Judith Ludwig and Christian Merten (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-12933.txt.
Introduction
Judith Ludwig and Christian Merten report a formal proof build in a 2025 paper. Their target is the Bruhat-Tits tree, a key tool in modern number theory that past work used on paper. The pair rebuild it inside the Lean theorem prover, where a machine checks each step. To prove the build can serve live research, they use it to verify a known result about harmonic cochains on the tree. The note is short, but it ties proof software to an active line of number theory work.
Catalog record
| CREATOR | Judith Ludwig and Christian Merten |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-12933
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Formalising the Bruhat–Tits Tree Judith Ludwig and Christian Merten Abstract In this article we describe the formalisation of the Bruhat–Tits tree—an important tool in modern number theory—in the Lean Theorem Prover. Motivated by the goal of connecting to ongoing research, we apply our formalisation to verify a result about harmonic cochains on the tree. keywords Lean, Bruhat–Tits tree, Cartan decomposition † † firstpage: 55 \authorinfo [J. Ludwig]Heidelberg [email protected] \authorinfo [C. Merten]Utrecht [email protected] \VOLUME 2 \YEAR 2026 \NUMBER 3 \DOI https://doi.org/10.46298/afm.15738 \receiveddate May 23, 2025 \finaldate February 18, 2026 \accepteddate April 14, 2026 \msc 68V20, 20E08, 20G25 1 Introduction The Bruhat–Tits tree is a combinatorial object that serves as a powerful tool in number theory and arithmetic geometry. In the simplest setting of the field \mathbb{Q}_{p} of p -adic numbers, the Bruhat–Tits tree is constructed from lattices in \mathbb{Q}_{p}^{2} and is a regular infinite tree, where each vertex has p+1 neighbours. Its geometry is characterised by its simplicity. However its connections to the structure theory and the representation theory of the group \mathrm{GL}_{2}(\mathbb{Q}_{p}) are deep and intricate. In this article we report on the formalisation of this useful mathematical concept in the Lean Theorem Prover. Our formalisation adds to the ongoing effort of formalising graduate level concepts in pure mathematics and more specifically in number theory. For the formalisation of the Bruhat–Tits tree, we pass to a more general setting and replace \mathbb{Q}_{p} by any discrete valuation field K . Before we describe our project in more detail, let us say a few more words on the applications of the Bruhat–Tits tree. For that it is useful to furthermore assume that K is complete and has finite residue field. In this setting, the Bruhat–Tits tree is a powerful tool for the study of subgroups of \mathrm{GL}_{2}(K) (as well as the closely related groups \mathrm{SL}_{2}(K) and \mathrm{PGL}_{2}(K) ) and their homology, see e.g. [ 10 , Chapter II] . Prominently, as explained in op. cit. it can be used to show Ihara’s theorem, that every torsion free subgroup of \mathrm{SL}_{2}(K) is free. The Bruhat–Tits tree is important in other parts of number theory, arithmetic geometry and representation theory, e.g., via its connection to Drinfeld’s upper half plane \Omega (see [ 5 ] ) or through the relation to classical as well as function field modular forms via harmonic cochains (see e.g. [ 12 , 11 ] ). The Bruhat–Tits tree has also been explored in mathematical physics (see e.g. [ 16 , 4 ] ). Bruhat–Tits theory puts the Bruhat–Tits tree into a larger number theoretic context. Keeping the setting of a non-archi-medean local field K for simplicity, let us remark that \mathrm{GL}_{2}/K is an example of a connected reductive group. Now much more generally, to any connected reductive group G/K one can associate a so called Bruhat–Tits building, a contractible topological space that carries the structure of a polysimplicial complex equipped with an action of G(K) . Bruhat–Tits buildings are a key tool for the structure theory of reductive groups, in spirit analogous to symmetric spaces of real reductive Lie groups. We refer to the introduction of [ 7 ] for an overview of the theory. The simplest non-trivial example is the building (as in [ 7 , Section 4.1] ) for G=\mathrm{GL}_{2} , where one recovers the Bruhat–Tits tree. Figure 1 is an illustration of the Bruhat–Tits tree for \mathbb{Q}_{2} , the field of 2 -adic numbers. Figure 1: The Bruhat–Tits tree for \mathbb{Q}_{2} We review the construction of the Bruhat–Tits tree and describe its formalisation in detail in Section 3 . For now let us mention that it is constructed as a simple graph, i.e., a graph such that any two vertices are connected by at most one edge, and such that no edge connects a vertex to itself; it has no loops. Moreover, it is connected and has no cycles making it into a so-called tree. As the edges out of any vertex are in bijection with points in the projective line over the residue field, the graph is regular, i.e., the edges out of any vertex have the same cardinality. The vertices of the Bruhat–Tits tree correspond to certain equivalence classes of lattices in K^{2} . To define the edges of this graph and to understand that it is indeed a tree, it is necessary to relate lattices to each other, i.e., to develop a concept of distance between lattices and to understand chains of lattices. This is achieved by studying how bases of lattices can be transformed into standard shapes and for that one uses a decomposition of the group \mathrm{GL}_{2}(K) called the Cartan decomposition . This decomposition exists for \mathrm{GL}_{n}(K) (and more generally for connected reductive groups) and is of independent interest. In the context of p -adic fields for example, it has consequences for the theory of smooth representations of the p -adic group \mathrm{GL}_{n}(K) . In Section 2 we give the precise statement, more details on context and describe the formalisation of this decomposition. Our formalisation adds to the ongoing formalisation effort of modern number theory. It is moreover motivated by connecting to current research. An ongoing research project of one of us (J.L.) prominently features the Bruhat–Tits tree via harmonic cochains. These are certain functions on the edges of the tree connecting to representation theory of \mathrm{GL}_{2} and to rigid analytic functions on Drinfeld’s period domain. As we strive to write up our research findings in a way that is error-free, accessible, understandable and unambiguous, we are intrigued by the idea that one day formalisation and verification tools will effortlessly support this process and improve on our documentation. This is still Zukunftsmusik , particularly when using the word effortless. But in an attempt to turn up the volume we applied our formalisation of the Bruhat–Tits tree to verify a result about harmonic cochains and gained some mathematical insights along the way. This application is explained in Section 4 below. We refer to Section 5.2 for further discussion on the research context. To our knowledge, the Bruhat–Tits tree has never been formalised before in a proof assistant. The code for this project is available at the public Github repository . We organised our roughly 8000 lines of code into five folders. The folder Graph contains the main constructions regarding the Bruhat–Tits tree, with the file Graph / Tree . lean containing the main theorem theorem BTtree . We refer to the Readme for a detailed overview of the structure of the repository. We accompany our explanations of the formalisation by some Lean code snippets. We have edited some of them for clarity of exposition. All links to declarations in mathlib or in our repository are pinned to a specific commit and will therefore remain valid. 1.1 Lean and mathlib This formalisation project was done in the interactive theorem prover Lean 4, which is also a functional programming language and is based on the Calculus of Inductive Constructions. For more background on Lean, we refer to [ 9 , 1 ] . We use the stable version v4 .19.0 . Our project is based on Lean’s monolithic mathematical library mathlib4 , an extensive and fast growing open source library of mathematics [ 13 ]