Twisted aughts of alternating involutions
Raghavendra N. Bhat, Cristian Cobeli, Shuta Iwai, Zimeng Ye, Alexandru Zaharescu · 2025
Read Twisted aughts of alternating involutions by Raghavendra N. Bhat, Cristian Cobeli, Shuta Iwai, Zimeng Ye, Alexandru Zaharescu (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-2509-17838.txt.
Introduction
Raghavendra N Bhat and 4 co-authors study a group built from sign-flip matrices with a twist. Their 2025 paper defines a subgroup of matrices over the whole numbers. Each key matrix matches the plain identity, save for one line where minus one and plus one take turns.
The team works out the traits of that group in depth. They then track what comes from using a fixed string of such flips again and again on a grid point. Paths from that loop show several clear traits. The work ties a simple flip rule to rich group behaviour.
Catalog record
| CREATOR | Raghavendra N. Bhat, Cristian Cobeli, Shuta Iwai, Zimeng Ye, Alexandru Zaharescu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 93 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-17838
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Twisted aughts of alternating involutions Thanks: Key words and phrases: involutions, twisted orbits, involutive matrices, orbit diameter, Coxeter groups Raghavendra N. Bhat, Cristian Cobeli, Shuta Iwai, Zimeng Ye, Alexandru Zaharescu Abstract . Let ℳ ( n ) \mathcal{M}(n) be the subgroup of G L ( n , ℤ ) GL(n,\mathbb{Z}) generated by the particular involutions that are identical to the identity, except for a single line where − 1 -1 and + 1 +1 alternate. We study the properties of ℳ ( n ) \mathcal{M}(n) , and then find several notable characteristics of the unions of trajectories obtained by iteratively applying a fixed sequence of such involutions to elements from ℤ n \mathbb{Z}^{n} . 2020 Mathematics Subject Classification Primary 11B37; Secondary 51F15, 11B50 1. Introduction In the quest for distinguished patterns in tiling problems or partitioning the space with integers, certain involutions arise and prove to have a directive role. Depending on their destined play, they can function in a lower dimension, as in [ 1 ] , where they were used to indicate location and content by changes in direction and size, or by keeping the dimension as in [ 5 , 4 ] , where they also take on the property of being invariant under translations. Within the same class, the type of operators we are considering are elements of GL(n,\mathbb{Z}) , which act on elements of \mathbb{Z}^{n} by replacing one component with an alternating sum of all components to which an additive constant is added while the other components remain unchanged. Nonetheless, that constant added to the alternating sum entangles a straightforward generalization of these transformations to higher dimensions. In this case, by selecting a feasible option, we have opted to put the constant aside, along with the requirement of invariance under translations, keeping only the one of being involutions. Moreover, in the most basic considered version, we can replace \mathbb{Z} with \mathbb{F}_{2} , but to emphasize the reflective property, we will keep the entries in \{-1,0,1\} instead. Let n be a positive integer. For any integer j , with 1\leq j\leq n , we let K_{j}=K_{j}(n) denote the modified identity matrix such that the j -th line starts with (-1)^{j} and then alternates \pm 1 ’s. To be precise, let \bm{e}_{j}=\bm{e}_{j}(n) be the j -th standard basis (column) vector and let \bm{r}_{j}=\bm{r}_{j}(n):=\big((-1)^{j},(-1)^{j+1},\dots,(-1)^{j+n-1}\big) . (We will omit the reference to dimension n as the dimension will be clear from the context.) Thus \displaystyle K_{j}:=Id-\bm{e}_{j}\bm{e}_{j}^{T}+\bm{e}_{j}\bm{r}_{j}. (1) We also consider the corresponding operators \K_{j}:\mathbb{Z}^{n}\to\mathbb{Z}^{n} that transform any vector \bm{x}=(x_{1},\dots,x_{n})\in\mathbb{Z}^{n} into the product (K_{j}\bm{x}^{T})^{T} . (Further on, in this context, we will drop the indication of transpositions by writing simply \K_{j}(\bm{x}) or just K_{j}\bm{x} .) Iteratively applying operators \K_{j} , with indices selected from a given sequence of positive integers, creates a subset of \mathbb{Z}^{n} , called trajectory or orbit , when it is closed. These trajectories vary based on the chosen initial point, called trajectory generator . In the case of non-trivial sequences, excluding generators belonging to certain special locations that are part of sets with zero measure in the working environment (such as the origin, specific axes, planes, or hyperplanes), the trajectories share the same class of shape indicated by the same underlying topology. For any fixed sequence of indices of the operators, the union of the associated trajectories forms a set partition of \mathbb{Z}^{n} . The mentioned additive constant, which is 0 in the case of operators \K_{j} , causes the trajectories generated by their iterated application to be bounded. This is in contrast to, for instance, the trajectories that were unbounded parallel parabolas or intricately looking circles that partition the plane \mathbb{Z}^{2} in [ 5 , 4 ] . Figure 1. Let \bm{w}=(10,8,15) and \K=\left(\K_{3}(\K_{2}\K_{1})^{2}\K_{2}\K_{3}(\K_{1}\K_{2})^{2}\K_{1}\right)^{2} . The image on the left-side shows the 24 nodes and 36 edges of the graph obtained by starting from \bm{w} and applying recursively any of the operators \K_{1},\K_{2},\K_{3} . The Euclidean diameters of the graph and their midpoints are also included. The Eulerian-Hamiltonian cycle, which connects the nodes of the orbit o(\bm{w}) in the order given by their appearance while applying the operators in \K , is shown in the middle image. In the right-hand image, one can see the projections onto the unit sphere of all the nodes of the orbits generated likewise by \K while starting from every \bm{x}\in[0,10]^{3} . (In the electronic version, the diametrically opposite nodes of the orbits they belong to are in red and the others are in blue.) Applying the K_{j} operators iteratively on points from \mathbb{Z}^{n} , the ubiquitous geometric shape that arises in all pairs of dimensions is that of a twisted aught , resembling that of tilted figure 8 , with lines drawn parallel to the coordinate axes. (See their algebraic expressions and graphic representations in Sections 7.1 and 9.3 .) In Figure 1 , one can see the graph with 24 nodes and 36 edges that contains any possible trajectory starting from (10,8,15) . The nodes are positioned in 12 levels, with 4 parallel to each of the three planes x_{j}=0 for j=1,2,3 , respectively. In the orbit shown in the center of Figure 1 , even though there are 6 nodes for each aught on each level, some edges are left untraced when jumping from one level to another, thus completing a Hamiltonian path, which is also an Eulerian cycle. There are orbits that also have notable arithmetic properties such as those related to squares and primes. Thus, an orbit that passes through a point with coordinates (a^{2},c^{2}) , where a^{2}+b^{2}=c^{2} for some integers a,b,c , contains only nodes whose coordinates are squares or their negatives; and an orbit that passes through (p,p+2) , where both p and p+2 are primes, has only prime numbers or their opposites in all components of the other nodes. A wide range of topics regarding problems related to groups generated by involutions and matters on the associated Hamiltonian cycles have been studied in many works; see for example [ 14 , 17 , 3 , 16 , 11 , 6 , 9 ] , the surveys [ 15 , 13 , 12 ] , and the references therein. 1.1. Main results Our work has two main parts. In Sections 2 – 5 , starting from the observation that all K_{j} with j=1,2,\dots,n are involutions, we follow step by step on their properties and arrive at a comprehensive description of the group they generate (see Theorem 6 ). Lately, we find that, as a Coxeter group, \langle K_{1},K_{2},\dots,K_{n}\rangle is isomorphic with the symmetric group S_{n+1} (see Section 5 ). In the second part, we turn our attention specifically to the characteristic aspects of sets formed by unions of orbits, particularly in the two-dimensional case. This proves to be essential for understanding the general phenomenon, which will be the topic of a subsequent work. Starting from an arbitrary \bm{x}\in\mathbb{Z}^{2} and applying the K_{j} operators iteratively, we obtain a cyclic trajectory o(\bm{x}) with 6 points, except for a few degenerate situations. In each orbit o(\bm{x}) , let’s color in red the points farthest from each other, that is, those that are diametrically opposed in Euclidean distance. We color the points that