Extending bounds on minimal ranks of universal quadratic lattices to larger number fields
Matěj Doležálek · 2025
Read Extending bounds on minimal ranks of universal quadratic lattices to larger number fields by Matěj Doležálek (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-2507-23338.txt.
Introduction
Matěj Doležálek pushes a number theory result about universal lattices into far larger fields. A universal quadratic lattice can represent every positive element of its number field, and some fields force such lattices to have huge minimal rank. Prior work built those examples in modest degrees, and Kala's technique stretched them to bigger fields under set conditions.
Doležálek sharpens the stretching tool by studying how subfields sit inside fields glued together from several pieces. Basic Galois theory turns that structure question into a problem about groups. His headline result is a clean lifting rule. If fields of degree d hold universal lattices of rank at least r, then fields of degree three times d, or any larger multiple, hold them too.
Catalog record
| CREATOR | Matěj Doležálek |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-23338
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Extending bounds on minimal ranks of universal quadratic lattices to larger number fields Thanks: The author acknowledges support by Charles University project GAUK No. 134824. Matěj Doležálek Abstract . There exist numerous results in the literature proving that within certain families of totally real number fields, the minimal rank of a universal quadratic lattice over such a field can be arbitrarily large. Kala introduced a technique of extending such results to larger fields – e.g. from quadratic fields to fields of arbitrary even degree – under some conditions. We present improvements to this technique by investigating the structure of subfields within composita of number fields, using basic Galois theory to translate this into a group-theoretic problem. In particular, we show that if totally real number fields with minimal rank of a universal lattice ≥ r \geq r exist in degree d d , then they also exist in degree k d kd for all k ≥ 3 k\geq 3 . 2020 Mathematics Subject Classification Primary 11E12; Secondary 11E20, 11R32, 11R80 1. Introduction The quest to understand quadratic forms spans much of the history of number theory. This started with early results of Fermat, Lagrange and Legendre, who studied the expressibility of positive integers as sums of two, four and three squares respectively, and continued throughout the following centuries. A particular interest has been directed at universal quadratic forms, i.e. positive definite forms that represent all positive integers. This culminated with Bhargava and Hanke [ BH ] finding all universal quaternary forms using their 290-theorem, which surprisingly states that a quadratic form is universal if and only if it represents all 29 of the so-called critical integers , the largest of which is 290 . While this provides a satisfying resolution over \mathbb{Z} , the situation is much less well-understood over number fields. Here, the usual setting is to consider a totally real number field and a form which attains totally positive values outside of the origin. A universal form is then one which represents all totally positive algebraic integers of the field. One may also slightly generalize quadratic forms to quadratic lattices , this essentially corresponds to considering arbitrary finitely generated modules equipped with quadratic forms instead of only free modules (see Subsection 2.1 for a precise definition). Some of the early ventures into quadratic forms over number fields include the study of sums of squares over \mathbb{Q}(\sqrt{5}) by Kirmse [ Ki ] and Götzky [ Go ] , over \mathbb{Q}(\sqrt{2}) and \mathbb{Q}(\sqrt{3}) by Cohn [ Ch ] and over arbitrary totally real number fields by Siegel [ Si ] , who proved that \mathbb{Q} and \mathbb{Q}(\sqrt{5}) are the only totally real fields where a sum of squares is universal. Surprisingly, while \mathbb{Q} requires four squares to express every positive integer, over \mathbb{Q}(\sqrt{5}) already the sum of three squares is universal (this is due to Maaß [ Ma ] ). Related to this, the currently still open Kitaoka’s conjecture posits that there are only finitely many totally real number fields which admit a ternary universal form. In recent years, indecomposable elements have been used extensively with regards to representation of algebraic integers by quadratic forms over totally real number fields: a totally positive algebraic integer is called indecomposable if it may not be written as a sum of two totally positive algebraic integers. These have been use to prove partial results towards Kitaoka’s conjecture or to determine which number fields admit a universal form with coefficients from \mathbb{Z} (e.g. [ KY1 , KY2 , KTZ ] ). For further background on techniques related to indecomposable elements as well as on the state of the art regarding Kitaoka’s conjecture, we refer the reader to a survey paper by Kala [ Ka3 ] . The strand of recent results most pertinent to the present article are the ones which construct fields in a chosen degree such that all universal forms over such a field must have a high number of variables, or more generally such that a universal quadratic lattice must have high rank. Denoting by m(F) the minimal rank of a universal lattice over a totally real number field F , fields with m(F)\geq r for arbitrary r are known to exist among quadratic fields [ BK , Ka1 ] , cubic fields [ KT , Ya ] , biquadratic [ CL+ ] and multiquadratic [ KS ] fields. In fact, later results of Kala, Yatsyna and Żmija [ KYZ ] and Man [ Mn ] respectively showed that almost all (in a sense of natural density) totally real quadratic and multiquadratic fields have m(F) larger than a given r . Building on this, Kala [ Ka2 ] introduced a technique for extending these results to higher the degrees. Starting from a suitable field L with a high m(L) , one finds a field K such that the structure of subfields in the compositum KL is “as simple as possible” (see Subsection 3.1 for a precise statement). Then, if \disc_{K} is sufficiently large, it can be shown that all universal lattices over KL in fact contain a universal lattice over L , which yields m(KL)\geq m(L) , hence m(KL) must also be large. With this, the results cited in the previous paragraph may be used to provide existence of totally real fields F with high m(F) in all degrees divisible by 2 or 3 . The original proof in [ Ka2 ] contains several technical assumptions, such as L being Galois over \mathbb{Q} with \Gal(L/\mathbb{Q}) cyclic or the Galois closure \tilde{K} of K over \mathbb{Q} having \Gal(\tilde{K}/\mathbb{Q})\simeq S_{[K:\mathbb{Q}]} . Our purpose in this paper is to treat the technique more generally, remove or ease these technical assumptions as much as possible and, in doing so, make it readily applicable if new “source results” appear in the future (e.g. a construction of F ’s with high m(F) in degree 5 , which our technique would immediately extend to all multiples of 5 ) – this is contained Corollary 4.3 . Our version of the main result is as follows (cf. [ Ka2 , Theorem 4] ): Theorem 1.1 . Let k , \ell be positive integers and L a totally real number field of degree \ell . Then for all totally real number fields K of degree k satisfying (i) \disc_{K} is greater than C(k,L) and coprime to \disc_{L} , (ii) K has no proper subfields apart from \mathbb{Q} , (iii) k\nmid\ell or K/\mathbb{Q} is not Galois, we have m(KL)\geq m(L) , where C(k,L) is a real constant dependent only on k and L . As a consequence, for all k\geq 3 there are infinitely many totally real number fields F of degree k\ell with m(F)\geq m(L) . We will prove this in Section 4 as Theorem 4.2 and Corollary 4.3 . 2. Preliminaries 2.1. Number fields and quadratic lattices A number field is a field extension F of the rational numbers \mathbb{Q} of some finite degree d and we denote its rings of algebraic integers as \mathcal{O}_{F} . Such an F is equipped with d embeddings \sigma_{1},\dots,\sigma_{d}:F\to\mathbb{C} ; when all their images lie in \mathbb{R} , we say F is totally real . In that case we say an \alpha\in