Density of integral points in the Betti moduli of quasi-projective varieties
Simone Coccia and Daniel Litt · 2025
Read Density of integral points in the Betti moduli of quasi-projective varieties by Simone Coccia and Daniel Litt (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-2507-00167.txt.
Introduction
Simone Coccia and Daniel Litt prove a density result for integral points in character varieties, in a math paper from 2025. The setting is a smooth variety with a simple normal crossings compactification. The varieties parametrize SL2 local systems whose boundary traces are fixed algebraic integers. The authors show integral points are potentially dense there. Their proof uses work of Corlette and Simpson to cut the claim down to Riemann surfaces. There, they build an integral point whose orbit under the mapping class group is Zariski dense.
Catalog record
| CREATOR | Simone Coccia and Daniel Litt |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 71 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-00167
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Density of integral points in the Betti moduli of quasi-projective varieties Simone Coccia and Daniel Litt Abstract . Let Y Y be a smooth quasi-projective complex variety equipped with a simple normal crossings compactification. We show that integral points are potentially dense in the (relative) character varieties parametrizing SL 2 \SL_{2} -local systems on Y Y with fixed algebraic integer traces along the boundary components. The proof proceeds by using work of Corlette-Simpson to reduce to the case of Riemann surfaces, where we produce an integral point with Zariski-dense orbit under the mapping class group. 1. Introduction 1.1. Conjecture and main result Let Y be a smooth complex variety equipped with a smooth projective simple normal crossings compactification \overline{Y} , with D=\overline{Y}\setminus Y . Given a commutative ring R and an affine algebraic group G/R , the G -representation variety \operatorname{Hom}(\pi_{1}(Y),G) is the affine R -scheme whose S -points for an R -algebra S are \operatorname{Hom}(\pi_{1}(Y),G)(S):=\operatorname{Hom}(\pi_{1}(Y),G(S)). The G -character variety of Y is the (categorical) quotient X_{G}(Y):=\operatorname{Hom}(\pi_{1}(Y),G)/G, where G acts by conjugation. For each component D_{i} of D , i=1,\cdots,n , fix a small loop \gamma_{i} around D_{i} and an R -point C_{i} in the adjoint quotient (G/_{\text{ad}}G)(R) . There is a natural map p_{D}:X_{G}(Y)\to(G/_{\text{ad}}G)^{n} induced by the map \rho\mapsto(\rho(\gamma_{i}))_{i=1,\cdots,n}. Setting \underline{C}=(C_{i})_{i=1,\cdots,n} we define the relative character variety X_{G,\underline{C}}(Y):=p_{D}^{-1}(\underline{C}). For example, if G=\SL_{2} , X_{G,\underline{C}}(Y) parametrizes conjugacy classes of representations \rho of \pi_{1}(Y) into \SL_{2} with \operatorname{tr}(\rho(\gamma_{i})) fixed. The goal of this paper is to provide some evidence for the following conjecture. Conjecture 1.1.1 . Let G be a Chevalley group over \mathbb{Z} , K a number field, and \mathscr{O}_{K} the ring of integers of K . Fix \underline{C}\in(G/_{\text{ad}}G)(\mathscr{O}_{K})^{n} . Then integral points are potentially Zariski-dense in the K -scheme X_{G,\underline{C}}(Y)_{K}. That is, there exists a finite extension L/K such that the Zariski-closure of the \mathscr{O}_{L} -points of X_{G,\underline{C}}(Y) contains X_{G,\underline{C}}(Y)_{K}. Recall that here a Chevalley group is a smooth affine group scheme over \mathbb{Z} with connected reductive fibers, admitting a fiberwise maximal \mathbb{Z} -torus; for example, \SL_{n,\mathbb{Z}},\PGL_{n,\mathbb{Z}} , \operatorname{Sp}_{2n,\mathbb{Z}} are Chevalley. Some version of this conjecture has been considered by a number of people; for example see [ Lit24 , Question 5.4.3(2)] . In this paper we verify 1.1.1 for G=\SL_{2},\PGL_{2} : Theorem 1.1.2 . Let G=\SL_{2,\mathbb{Z}} or \PGL_{2,\mathbb{Z}} . Fix a number field K and \underline{C}\in(G/_{\text{ad}}G)(\mathscr{O}_{K})^{n}. Then integral points are potentially dense in X_{G,\underline{C}}(Y). The proof proceeds by reduction to the case of curves, relying on Corlette-Simpson’s and Loray-Pereira-Touzet’s classification of rank 2 local systems on quasi-projective varieties [ CS08 , LPT16 ] . We handle the case where Y is a curve (say, of genus g with n\geq 0 punctures) by constructing, for every \underline{C} , an integral representation whose orbit under the pure mapping class group of a surface of genus g with n punctures is Zariski-dense in X_{G,\underline{C}}(Y) . As the action of the mapping class group preserves integrality, this suffices. Our results on Zariski-density of integral points in (relative) character varieties of surface groups are Theorem 5.0.4 and Theorem 6.0.1 . In particular, we show that if K is the field of definition of \underline{C} , then there exists a degree 4 extension L/K such that \mathscr{O}_{L} -points are Zariski-dense in the relative \SL_{2} -character variety of a curve of genus g with n punctures; see 5.0.5 . Some such field extension is necessary; see 5.0.6 . 1.2. Motivation and related work The primary antecedent to 1.1.1 is Simpson’s conjecture on integrality of rigid local systems [ Sim92 , p. 9] , which is precisely the statement that integral points are Zariski-dense in 0 -dimensional components of X_{G,\underline{C}}(Y) , at least when Y is projective. Even this case and its quasi-projective variant is open, though beautiful work of Esnault-Groechenig (in the case G=GL_{n} ) [ EG18 ] and Klevdal-Patrikis (for general G ) [ KP22 ] prove that reduced isolated points of X_{G,\underline{C}}(Y) are integral, for \underline{C} quasi-unipotent. De Jong-Esnault [ DJE24 ] show that, if non-empty, X_{G}(Y) has a \overline{\mathbb{Z}_{\ell}} -point for every \ell (and much more); this would evidently be a consequence of 1.1.1 . Under mild hypotheses they show X_{G}(Y) has a \overline{\mathbb{Z}} -point. All of these results rely on the existence of \ell -adic companions , due to Lafforgue [ Laf02 ] in dimension one and Drinfeld [ Dri12 ] in general, ultimately relying on Lafforgue’s work on the Langlands program for function fields over finite fields. Why might one believe 1.1.1 ? Aside from the fact that it generalizes Simpson’s conjecture to positive-dimensional components of X_{G,\underline{C}}(Y) , it is also motivated by a conjecture of Campana [ Cam11 , Conjecture 13.23 ] predicting which varieties should have a potentially Zariski-dense set of S -integral points. One particular instance of such conjectures is that log Calabi-Yau varieties 1 1 1 A variety Z is log Calabi-Yau if it admits a normal projective compactification X with reduced boundary divisor D such that K_{X}+D\sim 0 . admit an integral model with a Zariski-dense set of integral points (see [ Cam11 , Théorème 7.7 ] ). An expectation attributed to Kontsevich-Soibelman is that in many cases character varieties are “cluster varieties”, hence log Calabi-Yau (see the discussion after Conjecture 5 of [ KNPS15 ] ). Campana’s conjecture then predicts that they should have a Zariski dense set of integral points. Whang [ Wha20b , Theorem 1.1] has proven that relative \SL_{2} -character varieties of surfaces are log Calabi-Yau, so that our Theorem 5.0.4 answers positively Campana’s conjecture for such varieties. We remark that our result is stronger than the expectation of Campana’s conjecture, as we prove potential density of integral points rather than S -integral ones. Our results on Zariski-density of integral points for surface groups are closely related to the study of mapping class group dynamics on (relative) character varieties; indeed, we prove density by finding integral points with Zariski-dense mapping class group orbit. We rely on the study of the geometry of character varieties from [ Wha20a ] . Recently Golsefidy-Tamam [ AG ] (see also [ GT25 ] for a summary of results) have closely studied Zariski-density of mapping class group orbits in character varieties of surfaces; we expect we could have used their results for our purposes as well, though we have opted for a more self-contained exposition. In general, dynamics of mapping class groups on character varieties has been studied from a number of points of view by Goldman [ Gol05 , GX09 ] , Previte-Xia [ PX00 , PX02 ] , and others, including the second-named author and collaborators [ LLL23 , LL24 ] . Arithmetic aspects of \SL_{2} -character varieties of surfaces have recently been studied in the work of several authors. For instance, strong approximation results for surfaces of Markoff type (which are relative character varieties of the projective line with four punctures) have been established in the work of Bourgain-Gamburd-Sarnak [ BGS16 ] and Chen [ Che24 ] (see also [ Mar25 ] for a more elementary approach to part of Chen’s work). Ghosh and Sarnak [ GS22 ] investigated the integral Hasse principle for a family of Markoff cubic surfaces, showing (among various things) that almost all surfaces admitting a \mathbb{Z}_{p} solution at