HOW STARK UNITS ENTER SIC OVERLAPS

Ingemar Bengtsson*, Gary McConnell** · 2026

Read HOW STARK UNITS ENTER SIC OVERLAPS by Ingemar Bengtsson*, Gary McConnell** (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-2606-23535.txt.

Introduction

Bengtsson and McConnell trace a surprising bridge between quantum measurement theory and deep number theory, in a 2026 mathematical physics paper. A SIC is a highly symmetric set of quantum states whose pairwise overlaps, inner products measuring their alignment, take one fixed size. Those overlap values turn out to be algebraic units, special numbers built from Stark units, quantities that Stark's conjectures extract from zeta functions to generate certain number fields.

Mixing exact arguments with numerical evidence, the authors argue that overlap units are always products of powers of square roots of Stark units. In the non-minimal case, a whole lattice of ray class fields, layered number fields, enters the picture. In every second dimension, counted in a specific way, some overlap units collapse to plus or minus one. They trace this pattern to one special property of those ray class fields.

Catalog record

CREATORIngemar Bengtsson*, Gary McConnell**
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE118 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.23535

Browse the full library · About the Tower of Babel

Full text

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

HOW STARK UNITS ENTER SIC OVERLAPS Ingemar Bengtsson*, Gary McConnell** Abstract ∗ Stockholms Universitet, AlbaNova, SE-106 91 Stockholm, Sverige ∗∗ Controlled Quantum Dynamics Theory Group, Imperial College, London, United Kingdom SW7 2AZ Abstract: It has been observed that the mutual scalar products of the vectors in a SIC-POVM are given by algebraic units, and at least in some cases by square roots of Stark units. The full picture is somewhat more complicated, especially if non-minimal SIC-POVMs are considered. We present a mixture of exact and numerical evidence suggesting that the overlap units are always products of integral powers of square roots of Stark units from ray class fields all of which are attached to the maximal ring of integers in the base field. In the non-minimal case a lattice of such ray class fields is involved. In every second dimension (counted in a certain way) some of the overlap units equal \pm 1 , and we show that this follows from a special property of the ray class fields. Our observations are complementary to but consistent with the claim that the overlap units can be calculated directly from the Shintani–Faddeev modular cocycle. 1 Introduction We will be concerned with the intersection of two unsolved problems, one of them coming from number theory and the other from quantum theory. It came as a considerable surprise that this intersection exists [ 1 , 2 ] . The number theory problem is a subproblem of Hilbert’s 12th: find an elegant description of the generators of the most general abelian extension of a real quadratic number field. (This formulation is a little vague, but so was Hilbert’s [ 3 ] .) A real quadratic field is an extension K={\mathbb{Q}}(\sqrt{D}) of the rational field {\mathbb{Q}} , where D is a positive square-free integer. In general an extension of a number field is said to be abelian if its Galois group is abelian. It has been known for a long time what the abelian extensions are. One fixes an ideal in the ring of integers of the base field, in our case K , and defines what is known as a ray class field with that ideal as the finite part of its modulus. Every abelian extension is a subfield of such a ray class field. Fifty years ago Harold Stark gave a procedure that conjecturally allows us to calculate algebraic units in any such ray class field [ 4 ] . It starts from an analytic zeta function attached to the field and—provided that the numerical calculation is carried out with enough precision—ends with an exact expression for what are known as Stark units. In most but not quite all of the cases we consider they serve as generators of their ray class field. Much more can be said, a main point being that there does not at this time exist a proof of Stark’s conjectures, not even when the base field is K . A SIC—the name is a shortening of the unilluminating acronym SIC-POVM—can be projectively defined as an orbit of the Weyl–Heisenberg group in the Hilbert space {\mathbb{C}}^{d} with a certain equiangularity property [ 5 , 6 ] . We have placed the group theoretical details in Appendix A , so as to not prolong this introduction unnecessarily for readers in the know. Once these details are accepted it is enough to consider the group elements \{D_{i,j}\}_{i,j=0}^{d-1} , where D_{0,0} is the identity element. Choose a fiducial unit vector \lvert\Psi_{0}\rangle for the group to act on. By definition the orbit is a SIC if and only if there exist phase factors e^{i\theta_{j,k}} , the indices j and k not both zero, such that \langle\Psi_{0}\lvert D_{0,0}\lvert\Psi_{0}\rangle=1\ ,\hskip 22.76219pt\langle\Psi_{0}\lvert D_{j,k}\lvert\Psi_{0}\rangle=\frac{e^{i\theta_{j,k}}}{\sqrt{d+1}}\ . (1) By taking the absolute values we see that the d^{2} vectors in the orbit are equiangular, and it is easy to prove that more than d^{2} equiangular unit vectors cannot exist in dimension d . The open question is whether this upper bound can be achieved (in this way) for every choice of the dimension d . Good reasons to be interested in SICs come from foundational concerns in quantum theory and from some quantum engineering applications [ 7 ] , but we do not go into this here. We focus on the quite remarkable observation that the unspecified phase factors e^{i\theta_{j,k}} that appear in the definition seem to be algebraic units as soon as d>3 [ 2 ] . For the case d=p where p is a prime equal to 2 modulo 3 Gene Kopp established in four examples that they are indeed square roots of Stark units [ 8 ] . There are other cases where Stark units enter the overlaps in a slightly different way [ 9 , 10 ] . For brevity we will refer to the phase factors as overlap units . The bridge between Hilbert space and number theory is the remarkable formula [ 1 ] (d+1)(d-3)=f_{0}^{2}\Delta_{0}\ ,\hskip 14.22636pt\Delta_{0}=\left\{\begin{array}[]{ll}D&\mbox{if $D=1$ mod 4}\\ 4D&\mbox{otherwise.}\end{array}\right. (2) Here D is a square-free integer used to define a real quadratic field, \Delta_{0} is a fundamental discriminant , and d is the dimension of the Hilbert space in which the SIC exists. In our set-up, d is also the modulus defining a ray class field with the quadratic field as its base field. The prevailing conjecture, verified in all cases examined, is that this ray class field can be used to construct a SIC in dimension d [ 2 ] . The integer f_{0} is of interest too: its divisors allow us to define an entire lattice of abelian extensions of the minimal ray class field, and to predict the ‘spectroscopy’ of unitarily inequivalent SICs in that dimension [ 11 , 12 ] . Early studies of the SIC existence problem included extensive numerical searches, leading Scott and Grassl to develop a ‘SIC phenomenology’ that tells us what symmetries the SICs enjoy in what dimensions [ 13 , 14 ] . The full picture is quite intricate, but it can now be derived from the conjectures about what number fields are needed to construct the SICs. In a recent contribution Kopp has proposed a special function known as the Shintani–Faddeev modular cocycle [ 15 ] which is believed to produce the algebraic units that eventually, after the appropriate Galois transformation, appear as SIC overlaps, for every SIC [ 16 ] . The aim of the present paper is then to see exactly how these units relate to Stark units in standard ray class fields. We take it as given that for each divisor f of f_{0} in equation ( 2 ) there exists a SIC that can be constructed in a well defined subfield of a ray class field with (finite) modulus fd , and that there exists an algorithm that allows us to compute the Stark units \mbox{S}_{fd} in that ray class field. If f is itself composite, f=f_{1}\cdots f_{n} , there is a lattice of subfields each containing its own Stark units \mbox{S}_{f_{i}d} . We claim that for each choice of d>3 and f\lvert f_{0} there exists a SIC such that: • Every overlap unit is a product of square roots of Stark units \mbox{S}_{f_{i}d} . For non-minimal SICs

Read the complete text