On the trivial units property and the unique product property

Heiko Dietrich, Melissa Lee, André Nies, Marc Vinyals · 2026

Read On the trivial units property and the unique product property by Heiko Dietrich, Melissa Lee, André Nies, Marc Vinyals (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-2603-22640.txt.

Introduction

Dietrich, Lee, Nies, and Vinyals report computer experiments on group rings of torsion-free groups. They probe two traits: the trivial units property and the unique product property. Both link to Kaplansky's unit and zero-divisor conjectures. The team classifies certain symmetric nontrivial units in the binary group ring of the Hantzsche-Wendt group. That group was used in Gardam's disproof of the unit conjecture. They also present a new candidate group. It may fail one property yet keep the other, a split no group is yet known to show.

Catalog record

CREATORHeiko Dietrich, Melissa Lee, André Nies, Marc Vinyals
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE59398
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.22640

Browse the full library · About the Tower of Babel

Full text

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

On the trivial units property and the unique product property Thanks: The second author acknowledges the support of an Australian Research Council Discovery Early Career Researcher Award (project number DE230100579) The third and fourth authors acknowledge the support of the Marsden fund of New Zealand. The first author thanks the University of Auckland for the hospitality and support for various research stays. Heiko Dietrich, Melissa Lee, André Nies, Marc Vinyals Abstract . We report on some computational experiments related to the trivial units property and unique product property for group rings of torsion-free groups. These properties are related to Kaplansky’s unit and zero-divisor conjectures. Our investigations include a classification of certain symmetric non-trivial units in the binary group ring of the Hantzsche-Wendt group; this group was used in Gardam’s refutal of Kaplansky’s unit conjecture. We also exhibit and investigate a new candidate group that fails the unique units property but may satisfy the trivial unit property. No examples of groups with these properties are known to date. 1. Introduction Let L be a ring with identity, not necessarily commutative. A zero-divisor in L is a non-zero x\in L such that xy=0 or yx=0 for some nonzero y\in L . One says that x\in L is a unit if there exists y\in L such that xy=1 and yx=1 ; if L has no zero-divisors, then one of the conditions implies the other. We will be interested in the group ring R[G] of a group G and commutative ring R with identity. Our paper focusses on two interrelated conditions: the trivial units property for group rings, and the unique product property for groups. We begin by introducing these conditions and giving brief background. Definition 1.1 . A group G satisfies the trival units property (TUP) for an integral domain R if the only units in the group ring R[G] are of the form rg , where g\in G and r is a unit of R . Such units will be called trivial. The following long-open conjecture was refuted by Gardam [ 12 ] . Conjecture 1.2 . If G is a torsion-free group and R an integral domain, then G satisfies the TUP for R . Gardam [ 12 ] found a non-trivial unit in \mathbb{F}_{2}[P] , where \mathbb{F}_{2} is the field with two elements and P is the torsion-free group defined by (1.1) \displaystyle P \displaystyle=\langle a,b\mid b^{-1}a^{2}b=a^{-2},a^{-1}b^{2}a=b^{-2}\rangle. We note that P is a crystallographic group, and x=a^{2},y=b^{2} and z=abab (in the notation of [ 12 ] ) are free generators of the abelian translation subgroup of P , which has index 4 in P . Remark 1.3 . While the proof given in Gardam’s paper is purely mathematical, the original discovery of a non-trivial unit in \mathbb{F}_{2}[P] was made computationally using a SAT solver. Gardam found a non-trivial unit supported on a ball of radius 5 around the identity in the Cayley graph of P . To do so, he reformulated the problem of finding a unit as a Boolean satisfiability (SAT) problem, assigning Boolean variables for the coefficients in \mathbb{F}_{2} to words in the generators of P and their inverses. The following purely group-theoretic property was introduced by Rudin and Schneider [ 29 ] , who used the term \Omega -group. Definition 1.4 . A group G has the unique product property (UPP) if for all non-empty subsets A,B\subseteq G , there exist some a\in A and b\in B such that whenever ab=a^{\prime}b^{\prime} with a^{\prime}\in A , b^{\prime}\in B , then a=a^{\prime} and b=b^{\prime} . In other words, when viewing AB as a multiset, some element occurs with multiplicity 1 . Promislow [ 25 ] showed that P fails the UPP. It is well known that UPP implies the TUP for every domain; see Proposition 2.1 below for a proof. Suprisingly, the converse implication is still unknown: is there a group that satisfies the TUP for some domain, but fails the UPP? This question is implicit in early work, and explicitly asked for instance in [ 12 ] . This paper follows an experimental approach, using SAT solvers and computer algebra towards new examples, which are then evaluated for advancing the theory. We • describe nontrivial units in \mathbb{F}_{2}[P] supported on balls from radius 4 onwards, and their potential symmetry properties; • expose a new candidate group that fails the UPP, but may satisfy the TUP for some domain. This group, denoted H_{4} , is a small index extension of the usual integral Heisenberg group. 2. Mathematical background 2.1. Group rings For a group G and a commutative ring R with multiplicative identity, the group ring R[G] consists of all the formal sums u=\sum_{g\in G}u_{g}g where only finitely many coefficients u_{g}\in R are non-zero. The addition operation is carried out component-wise. The operation of multiplication is given by (\sum_{g\in G}u_{g}g)(\sum_{h\in G}v_{h}h)=\sum_{r\in G}w_{r}r , where w_{r}=\sum_{gh=r}u_{g}v_{h} . By slight abuse of notation, we usually write 1=1_{G} and 1=1_{R} . The support of \sum_{g\in G}u_{g}g is the subset \{g\in G\mid u_{g}\neq 0\} . 2.2. Some background on the unit conjecture Higman posed Conjecture 1.2 in his unpublished 1940 PhD thesis [ 15 , p. 77] for the case that R={\mathbb{Z}} . The conjecture was taken up in generality by Kaplansky [ 18 ] , and became known as the unit conjecture . In his paper, Kaplansky attributes it to a list of problems arising from a 1968 conference in Moldova. Note that if R[G] has a non-trivial unit, then after extending the integral domain R to its field of fractions K , one obtains a counterexample in the group ring K[G] . Kaplansky also posed the stronger zero divisor conjecture: K[G] has no zero divisors, and the even stronger idempotent conjecture , that the only idempotents in K[G] are 0,1 . There are various families of groups that have the TUP for each integral domain R . Higman showed this for abelian groups, and more generally for locally indicable groups ; this means that every non-trivial finitely generated subgroup has a quotient isomorphic to \mathbb{Z} . After Gardam’s refutation [ 12 ] , his work was generalised by Murray [ 21 ] , who altered the construction to obtain non-trivial units in \mathbb{F}_{q}[P] for every prime q . Gardam [ 13 ] found a non-trivial unit in \mathbb{C}[P] ; it has the same support as the original example over \mathbb{F}_{2}[P] . Thus, the conjecture is refuted for fields of each characteristics. However, the original conjecture of Higman was for integral group rings, and remains open. 2.3. Background on the unique product property The study of the UPP defined in Definition 1.4 has a long history, and was instrumental to the refutation of the unit conjecture [ 12 ] . On the positive side, every bi-orderable group has UPP because the product of the maxima of two sets is unique; in fact, the weaker property of being diffuse suffices [ 2 ] ; for definitions and a proof see Proposition 2.1 below. The first example of a torsion-free group that fails the UPP was constructed by Rips and Segev [ 27 ] who made use of small cancellation theory. Passman introduced the torsion-free group P of ( 1.1 ) in connection with

Read the complete text