Prescribed lifts of 2-dimensional representations

Matthew Emerton, Toby Gee, Lue Pan, Xinwen Zhu · 2026

Read Prescribed lifts of 2-dimensional representations by Matthew Emerton, Toby Gee, Lue Pan, Xinwen Zhu (2026) 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-2606-25485.txt.

Introduction

Lifting a mod p representation of a Galois group to characteristic zero, with prescribed local behaviour, is a central engine behind modern modularity results. Matthew Emerton, Toby Gee, Lue Pan, and Xinwen Zhu prove a flexible lifting theorem in 2026 for two dimensional representations over a totally real field. Starting from an irreducible, totally odd residual representation, they allow the local conditions at each place to be any chosen union of components of the relevant deformation rings.

Away from the prime, any nonempty component choice is allowed. Above the prime, lifts may be potentially semistable with arbitrary regular Hodge Tate weights, a wide menu of local shapes. Under standard Taylor Wiles style hypotheses, plus extra care at the primes two and five, a global lift meeting every prescription exists. The result strengthens the local to global machinery that deformation theory runs on.

Catalog record

CREATORMatthew Emerton, Toby Gee, Lue Pan, Xinwen Zhu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.25485

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Prescribed lifts of 2-dimensional representations Thanks: ME was supported in part by the NSF grants DMS-1902307, DMS-1952705, DMS-2201242 and DMS-2502609. TG was supported in part by an ERC Advanced grant. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 884596). ME and TG were both supported in part by the Simons Collaboration on Perfection in Algebra, Geometry and Topology. LP was supported in part by a Sloan Research Fellowship. XZ was supported by NSF grant DMS-2200940. For the purpose of open access, the authors have applied a CC BY public copyright licence to any author accepted manuscript arising from this submission. Matthew Emerton, Toby Gee, Lue Pan, Xinwen Zhu Abstract . Let F F be a totally real field, and let p p be prime. Under standard Taylor–Wiles hypotheses, we show that an irreducible totally odd Galois representation ρ ¯ : Gal F → GL 2 ⁡ ( 𝐅 ¯ p ) \overline{\rho}:\Gal_{F}\to\GL_{2}(\overline{{\mathbf{F}}}_{p}) admits lifts lying on arbitrary prescribed components of local deformation rings, allowing potentially semistable conditions with arbitrary regular Hodge–Tate weights. 1. Introduction Let F be a totally real field, and let \Sigma be a finite set of finite places of F containing all places above a fixed prime p . We write \Gal_{F}\coloneq\Gal(\overline{F}/F) for the absolute Galois group of F , and we denote by \Gal_{F,\Sigma} the Galois group of the maximal extension of F unramified at all finite places outside \Sigma . We consider the following local-to-global lifting problem: let {\mathcal{O}} be the ring of integers in a finite extension of {\mathbf{Q}}_{p} , with residue field {\mathbf{F}} , and let \overline{\rho}:\Gal_{F,\Sigma}\to\GL_{2}({\mathbf{F}}) be a continuous representation. Suppose that we are given a de Rham character \psi:\Gal_{F,\Sigma}\to{\mathcal{O}}^{\times} lifting \det\overline{\rho} , and local conditions on lifts of \overline{\rho} at the places in \Sigma . At places away from p these conditions are allowed to be any non-empty union of irreducible components of the appropriate fixed-determinant local deformation ring, while at places above p they are allowed to be any non-empty union of irreducible components of a potentially semistable deformation ring with prescribed regular Hodge–Tate weights. We write {\mathcal{S}} for this collection of data, and R_{{\mathcal{S}}} for the corresponding global deformation ring; the precise definition is given in Definition 2.1.2 . Our main theorem is the following. Theorem 1.1.1 (Theorem 4.2.1 ) . Suppose that \overline{\rho} is totally odd, and that \overline{\rho}|_{\Gal_{F(\zeta_{p})}} is absolutely irreducible. If p=2 , assume that \overline{\rho} has non-solvable image, and if p=5 , assume that \overline{\rho} is not exceptional in the sense of Definition 4.1.1 . Then, for every global deformation problem {\mathcal{S}} , the ring R_{{\mathcal{S}}} is a finite {\mathcal{O}} -algebra of positive {\mathcal{O}} -rank. In particular, the set of characteristic-zero lifts of \overline{\rho} to representations of \Gal_{F,\Sigma} of type {\mathcal{S}} is finite and non-empty. In short, subject to the hypotheses of the theorem, arbitrary prescribed non-empty sets of local components occur globally. Remark 1.1.2 . Although the theorem as we have phrased it (allowing our local conditions to be unions of irreducible components of local deformation rings) is equivalent to the analogous version in which we prescribe a single choice of irreducible component at each prime, we find our phrasing convenient. For example, imposing a bound on the prime-to- p conductor of lifts of \overline{\rho} amounts to considering a certain union of local components. Thus the theorem immediately implies that the set of lifts of \overline{\rho} satisfying any given bound on the prime-to- p conductor (compatible with the prime-to- p conductor of \overline{\rho} ), as well as any prescribed p -adic Hodge-theoretic conditions at p , is finite and non-empty. In particular, the theorem gives a Galois-theoretic analogue of theorems on level raising and lowering, and on change of weight, for Hilbert modular forms; see [ 6 , §1.2] for a historical overview of this connection, an explanation of the basic strategy for proving such results (which goes back to Khare–Wintenberger), and an account of the subsequent generalisations to Hilbert modular forms (see for example [ 18 , 19 , 20 , 15 ] ). In brief: the global deformation ring with the prescribed local conditions has a Galois-cohomological lower bound on its dimension, which in the present setting says that \dim R_{{\mathcal{S}}}\geq 1 ; this is Proposition 2.1.4 . Thus, if one also knows that R_{{\mathcal{S}}} is finite over {\mathcal{O}} , this lower bound forces R_{{\mathcal{S}}} to have positive {\mathcal{O}} -rank, so that R_{{\mathcal{S}}}\otimes_{{\mathcal{O}}}\overline{{\mathbf{Q}}}_{p} is a non-zero finite \overline{{\mathbf{Q}}}_{p} -algebra. The points of its \Spec are exactly the desired lifts. The crux of the argument is therefore to prove the finiteness of the global deformation ring. In previous work, this finiteness is ultimately obtained from automorphy: after a suitable potential modularity argument, the deformation ring is compared with a Hecke algebra, which is finite over {\mathcal{O}} . For two-dimensional totally odd representations over totally real fields, the required finiteness in “parallel weight 2 ” is already known. Here this means that the labelled Hodge–Tate weights are \{0,1\} at every embedding; the terminology comes from the fact that the Galois representations associated to Hilbert modular forms of parallel weight 2 have these Hodge–Tate weights. In the form we need, this is a theorem of Snowden [ 27 ] (building on Kisin’s modularity lifting theorems [ 23 ] ), which is recalled below as Proposition 4.1.2 . The main new point of this paper is that we can deduce the general case of arbitrary regular Hodge–Tate weights from this weight 2 case. We do not use any new modularity lifting theorems to do this; indeed, we do not establish the potential modularity of our lifts. Instead, we make use of the geometry of the stack {\mathcal{X}} of étale (\varphi,\Gamma) -modules [ 12 ] (or, more precisely, its fixed-determinant variant). A “topological” form of the Breuil–Mézard philosophy, established in [ 12 ] , states that the underlying reduced substacks of the relevant potentially semistable substacks are equal to suitable unions of components of the underlying reduced substack {\mathcal{X}}_{\operatorname{red}} . Furthermore, each of these components already occurs in the underlying reduced substack of a potentially semistable substack of parallel weight 2 . (Indeed, Proposition 3.1.2 shows that every residual two-dimensional local representation with the required determinant is contained in the special fibre of a tamely potentially semistable weight 2 substack.) Passing from stacks to local deformation rings via versal maps gives Corollary 3.2.1 : modulo \varpi and up to nilpotent thickening, any potentially semistable local deformation ring is a quotient of a weight 2 local deformation ring. Tensoring these local comparisons together and using the standard description of global deformation rings as global-over-local tensor products then gives a surjection R_{{\mathcal{S}}^{\prime}}/\varpi\twoheadrightarrow R_{{\mathcal{S}}}/(\varpi,I), where {\mathcal{S}}^{\prime} is a weight 2 deformation problem and I is nilpotent. Since R_{{\mathcal{S}}^{\prime}} is finite over {\mathcal{O}} , it follows that R_{{\mathcal{S}}}/\varpi is finite over {\mathbf{F}} , and hence that R_{{\mathcal{S}}} is finite over {\mathcal{O}} . As sketched above, the Galois-cohomological lower bound then completes the proof of the theorem. Remark 1.1.3 . In the particular case when p splits completely in F , it is well-known that Theorem 1.1.1 can be deduced from the Breuil–Mézard conjecture for \operatorname{GL}_{2}/\mathbf{Q}_{p}

Read the complete text