Arithmetic Y-Frieze Patterns

• Arithmetic Y-frieze patterns are closed, staggered arrays of positive integers defined by a local Y-diamond relation and global glide symmetry.
• A key result is that every arithmetic Y-frieze arises from an SL₂-frieze via a surjective map rooted in cluster ensemble concepts.
• Finite classifications, such as 10 patterns for width 3 and 42 for width 4, highlight the combinatorial intricacies and arithmetic constraints of the theory.

Arithmetic Y-frieze patterns are closed Y-frieze patterns whose nonzero entries are all positive integers. In the type $A$ setting, a Y-frieze is a staggered infinite array of rational numbers governed locally by the Y-diamond relation

$WE=(1+N)(1+S),$

with boundary rows presented as rows of $0$'s, and with global glide symmetry and period $n+3$ in width $n$ (Germain, 2024). They were introduced as a coefficient-type analogue of Coxeter’s frieze patterns, motivated by $Y$-systems, cluster algebras, and cluster ensembles, and they now form a distinct arithmetic theory with finiteness theorems, explicit small-width classifications, and a complete realization theorem showing that every Y-frieze arises from an $\mathrm{SL}_2$-frieze (Germain, 2023, Short et al., 7 Jul 2026).

1. Definitions and basic local structure

A Y-frieze pattern is an infinite staggered array of rational numbers $(b_{i,j})$ in which every adjacent diamond

$$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$

satisfies

$WE=(1+N)(1+S).$

A Y-frieze is called closed if there is a row of zeros at the top and another row of zeros after finitely many nonzero rows, and the number of nonzero rows between these zero rows is called the width. A Y-frieze pattern is arithmetic if all its nonzero entries are positive integers; the set of arithmetic width-$WE=(1+N)(1+S),$0 Y-friezes is denoted $WE=(1+N)(1+S),$1 (Matsuzaki et al., 21 Aug 2025, Germain, 2024).

The local rule differs essentially from the Coxeter diamond rule. Classical Coxeter frieze patterns satisfy

$WE=(1+N)(1+S),$2

or, equivalently in the standard indexing,

$WE=(1+N)(1+S),$3

with boundary data involving rows of $WE=(1+N)(1+S),$4's. By contrast, Y-friezes replace the right-hand side by factors of the form $WE=(1+N)(1+S),$5, their first nonzero row is not forced to be all ones, and arithmeticity means positivity and integrality of all nonzero entries rather than positivity of a distinguished seed row alone (Germain, 2024, Germain, 2023).

The first nonzero row is called the zeroth row in one indexing convention, and the first nontrivial row is called the Y-quiddity cycle in the finite-type formulation. In width $WE=(1+N)(1+S),$6, glide symmetry takes the form

$WE=(1+N)(1+S),$7

and applying it twice yields periodicity by translation through $WE=(1+N)(1+S),$8 steps (Germain, 2024).

2. Cluster-algebraic and ensemble-theoretic origin

Y-frieze patterns were introduced as a variant of frieze patterns associated to acyclic cluster algebras. For a symmetrisable generalized Cartan matrix $WE=(1+N)(1+S),$9, a classical frieze pattern is replaced by a $0$0-frieze pattern satisfying a recurrence in which the classical exchange term $0$1 is shifted into products of $0$2-factors. In type $0$3, this reduces exactly to the local Y-diamond rule above, with top and bottom rows equal to $0$4 (Germain, 2023).

The conceptual source is the Fock–Goncharov ensemble

$0$5

where $0$6 is the $0$7-side cluster algebra, $0$8 is the $0$9-side positive space, and $n+3$0 is the ensemble map. The $n+3$1-variables mutate by the standard Fock–Goncharov rule, and the induced relations along a distinguished subtree of the cluster tree are exactly the Y-frieze recurrences. In this sense, Y-frieze patterns are the recursive shadow of $n+3$2-pattern mutation dynamics (Germain, 2023).

A central arithmetic subclass is formed by unitary Y-frieze patterns. For a chosen $n+3$3-cluster $n+3$4, one obtains a unique arithmetic Y-frieze

$n+3$5

Equivalently, a unitary Y-frieze is obtained by evaluating the universal Laurent expressions for the $n+3$6-variables at $n+3$7. The assignment from unordered $n+3$8-clusters to arithmetic Y-friezes is well-defined but generally neither injective nor surjective (Germain, 2023).

In rank $n+3$9, the theory becomes particularly explicit. For

$n$0

arithmetic Y-frieze patterns are exactly friezes of the corresponding generalized cluster algebra with recurrence

$n$1

The finite-type counts are

$n$2

and the paper proves a sharp criterion: finite type yields finitely many arithmetic Y-frieze patterns, while infinite type yields infinitely many (Germain, 2023).

3. Closed type $n$3 Y-friezes: knitting, symmetry, and arithmetic finiteness

A closed width-$n$4 Y-frieze can be studied constructively. From the Y-diamond rule, if $n$5 are known and $n$6, then

$n$7

This gives a vertical knitting process. However, not every Y-frieze can be generated in this way, because a row of $n$8's may appear and stop the recursion; examples include a case where an entire row becomes $n$9 (Germain, 2024).

For closed Y-friezes, horizontal knitting is cleaner. One places two rows of $Y$0's with $Y$1 empty rows between them, chooses $Y$2 positive rational numbers along a zig-zag path, and fills in the rest using

$Y$3

This yields a bijection between $Y$4-tuples of positive rational numbers and $Y$5-valued Y-frieze patterns of width $Y$6. Thus a width-$Y$7 Y-frieze is determined by the data along a chosen zig-zag (Germain, 2024).

Arithmeticity is substantially subtler than in the Coxeter case. The paper gives a simple arithmetic Y-frieze whose $Y$8-th row is constant and equal to

$Y$9

so the first few nonzero rows are $\mathrm{SL}_2$0. But positive integer initial data do not automatically produce an arithmetic Y-frieze: Y-frieze entries are generally not Laurent polynomials in the initial diagonal entries, and a width-$\mathrm{SL}_2$1 Y-frieze knitted from all $\mathrm{SL}_2$2's on a zig-zag produces entries such as

$\mathrm{SL}_2$3

This failure of automatic integrality is one of the main arithmetic distinctions from classical friezes (Germain, 2024).

The same paper proves a general finiteness theorem: for each fixed $\mathrm{SL}_2$4, the number of arithmetic Y-frieze patterns of width $\mathrm{SL}_2$5 is finite. It also records the complete lists for the smallest widths: $\mathrm{SL}_2$6 for width $\mathrm{SL}_2$7, and

$\mathrm{SL}_2$8

for width $\mathrm{SL}_2$9. For width $(b_{i,j})$0, it lists ten diagonals and conjectures that the list is complete, a conjecture later proved (Germain, 2024, Matsuzaki et al., 21 Aug 2025).

4. Complete classifications in widths $(b_{i,j})$1 and $(b_{i,j})$2

The classification problem has been solved completely for widths $(b_{i,j})$3 and $(b_{i,j})$4. In width $(b_{i,j})$5, a fundamental domain is determined by entries $(b_{i,j})$6 on the first diagonal, and the Y-diamond relations give

$(b_{i,j})$7

$(b_{i,j})$8

$(b_{i,j})$9

Arithmeticity requires all of these to be positive integers and leads to a finite inequality system (Matsuzaki et al., 21 Aug 2025).

The resulting classification is Theorem 10(abc): the triple $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$0 occurs as the first diagonal of an arithmetic Y-frieze of width $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$1 if and only if

$$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$2

Hence there are exactly $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$3 arithmetic Y-frieze patterns of width $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$4 (Matsuzaki et al., 21 Aug 2025).

In width $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$5, a fundamental domain is determined by $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$6, and the remaining entries $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$7 are given explicitly as rational functions of these variables. Integrality again produces a system of inequalities. Theorem 42(abcd) states that there are exactly $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$8 arithmetic Y-frieze patterns of width $$\begin{matrix} & N & \ W && E\ & S & \end{matrix}$$9, listed explicitly by $WE=(1+N)(1+S).$0 integer $WE=(1+N)(1+S).$1-tuples

$WE=(1+N)(1+S).$2

Thus the width-$WE=(1+N)(1+S).$3 arithmetic Y-friezes are completely classified and counted (Matsuzaki et al., 21 Aug 2025).

The proofs are driven by finiteness bounds. For width $WE=(1+N)(1+S).$4, Proposition 2.1 shows that if $WE=(1+N)(1+S).$5 and $WE=(1+N)(1+S).$6, then a necessary inequality fails, while Proposition 2.2 gives

$WE=(1+N)(1+S).$7

under $WE=(1+N)(1+S).$8, with the symmetric bound obtained by swapping $WE=(1+N)(1+S).$9 and $WE=(1+N)(1+S),$00. For width $WE=(1+N)(1+S),$01, the successive propositions force at least one of $WE=(1+N)(1+S),$02 to be small, then bound one of $WE=(1+N)(1+S),$03, then force $WE=(1+N)(1+S),$04, then bound one of $WE=(1+N)(1+S),$05 by $WE=(1+N)(1+S),$06, and finally show that under the remaining cases $WE=(1+N)(1+S),$07 and $WE=(1+N)(1+S),$08 are bounded by $WE=(1+N)(1+S),$09. The classification follows from a finite search within these explicit bounds (Matsuzaki et al., 21 Aug 2025).

These widths are special because the recurrence relations can still be controlled by hand plus finite computation, the arithmetic condition gives strong inequality constraints, and the number of possibilities remains finite and manageable (Matsuzaki et al., 21 Aug 2025).

5. The map $WE=(1+N)(1+S),$10 and realization by $WE=(1+N)(1+S),$11-friezes

A persistent theme in the subject is the relation between arithmetic Y-friezes and classical arithmetic friezes. De Saint Germain’s work defines a map

$WE=(1+N)(1+S),$12

and formulates the conjecture that $WE=(1+N)(1+S),$13 is surjective for all $WE=(1+N)(1+S),$14. In the type $WE=(1+N)(1+S),$15 closed setting, the Y-frieze is determined by the first two nontrivial diagonals of the classical frieze, and the construction is described as the frieze analogue of the cluster ensemble map of Fock and Goncharov (Germain, 2024).

The conjecture was verified first in small width. Since there are exactly $WE=(1+N)(1+S),$16 arithmetic Y-friezes of width $WE=(1+N)(1+S),$17, and they can be matched explicitly to Coxeter friezes of width $WE=(1+N)(1+S),$18, it follows that

$WE=(1+N)(1+S),$19

is surjective. For width $WE=(1+N)(1+S),$20, earlier work implies that $WE=(1+N)(1+S),$21 is injective for even $WE=(1+N)(1+S),$22, and because

$WE=(1+N)(1+S),$23

one obtains that

$WE=(1+N)(1+S),$24

is bijective (Matsuzaki et al., 21 Aug 2025).

The full conjecture was resolved by showing that all Y-friezes come from $WE=(1+N)(1+S),$25-friezes. For a staggered $WE=(1+N)(1+S),$26-frieze $WE=(1+N)(1+S),$27 of width $WE=(1+N)(1+S),$28, the associated Y-frieze is defined by

$WE=(1+N)(1+S),$29

If $WE=(1+N)(1+S),$30 satisfies the $WE=(1+N)(1+S),$31-diamond rule

$WE=(1+N)(1+S),$32

then $WE=(1+N)(1+S),$33 satisfies

$WE=(1+N)(1+S),$34

Theorem B proves that

$WE=(1+N)(1+S),$35

is surjective for every width $WE=(1+N)(1+S),$36 (Short et al., 7 Jul 2026).

The proof is constructive. Starting from a Y-frieze $WE=(1+N)(1+S),$37, the paper defines auxiliary quantities $WE=(1+N)(1+S),$38 and then integers $WE=(1+N)(1+S),$39 with

$WE=(1+N)(1+S),$40

A key lemma proves

$WE=(1+N)(1+S),$41

and an induction yields that each $WE=(1+N)(1+S),$42 is a positive integer. For even $WE=(1+N)(1+S),$43, these integers directly produce an integral $WE=(1+N)(1+S),$44-frieze, uniquely. For odd $WE=(1+N)(1+S),$45, one first obtains a rational $WE=(1+N)(1+S),$46-frieze and then rescales its quiddity to restore integrality (Short et al., 7 Jul 2026).

The fibers of $WE=(1+N)(1+S),$47 are extremely small. At most two $WE=(1+N)(1+S),$48-friezes map to the same Y-frieze. If two distinct friezes $WE=(1+N)(1+S),$49 map to the same Y-frieze, then $WE=(1+N)(1+S),$50 must be odd and their quiddities are related by alternating scaling,

$WE=(1+N)(1+S),$51

with

$WE=(1+N)(1+S),$52

The same ambiguity is characterized combinatorially by multiflips of triangulated polygons and geometrically by alternating rescaling of chains of tangent horocycles. In addition, the frieze entries admit hyperbolic interpretations: $WE=(1+N)(1+S),$53 where $WE=(1+N)(1+S),$54 are Ford circles and $WE=(1+N)(1+S),$55 are Farey edges (Short et al., 7 Jul 2026).

A frequent misconception in the earlier literature was that Y-friezes might form an independent class unrelated to classical $WE=(1+N)(1+S),$56-friezes. The surjectivity theorem shows that this is not the case: every Y-frieze is the product shadow of an $WE=(1+N)(1+S),$57-frieze (Short et al., 7 Jul 2026).

6. Relation to classical friezes and neighboring generalizations

Arithmetic Y-frieze patterns are best understood against the background of classical integer friezes. In the Coxeter–Conway theory, a frieze pattern consists of shifted rows beginning with a row of $WE=(1+N)(1+S),$58's and a row of $WE=(1+N)(1+S),$59's, every adjacent diamond satisfies

$WE=(1+N)(1+S),$60

and a closed integral frieze is determined by its quiddity sequence. The Conway–Coxeter theorem gives a bijection between frieze patterns of order $WE=(1+N)(1+S),$61 and triangulations of convex $WE=(1+N)(1+S),$62-gons, and triangulations correspond to clusters in type $WE=(1+N)(1+S),$63 cluster algebras (Baur, 2021). Arithmetic Y-friezes preserve the frieze paradigm of local exchange, periodicity, and finite arithmetic families, but their arithmetic is more delicate because positivity of initial data does not force positivity or integrality of the whole array (Germain, 2024).

There are also broader frieze generalizations that are adjacent to, but distinct from, arithmetic Y-friezes. One surface-theoretic generalization studies positive integral friezes on bordered marked surfaces with decorated hyperbolic metrics, where frieze entries are $WE=(1+N)(1+S),$64-lengths of arcs and the local rule is the Ptolemy relation

$WE=(1+N)(1+S),$65

For a pair of pants, every positive integral frieze is unitary, the unitary triangulation is unique, and short diagonals count crossed triangles in that triangulation (Canakci et al., 2021). This is not a Y-frieze theory, but it exhibits a parallel rigidity phenomenon in a different generalization of Conway–Coxeter friezes.

A second neighboring direction concerns arithmetic infinite friezes from triangulated punctured discs. There, periodic infinite friezes still satisfy the classical unimodular rule, but each south-east diagonal decomposes into arithmetic progressions, and the entries admit interpretations via matching numbers and a labeling algorithm on periodic triangulations of strips (Tschabold, 2015). This again shows that “arithmetic frieze” phenomena extend beyond the finite Coxeter case, although the governing local rule remains classical rather than Y-type.

The present state of the theory is therefore sharply differentiated. For arithmetic Y-frieze patterns of fixed width, finiteness is known in general; widths $WE=(1+N)(1+S),$66 and $WE=(1+N)(1+S),$67 are completely classified; and every Y-frieze is known to arise from an $WE=(1+N)(1+S),$68-frieze (Germain, 2024, Matsuzaki et al., 21 Aug 2025, Short et al., 7 Jul 2026). At the same time, no Catalan-type formula is known in general for $WE=(1+N)(1+S),$69, and the arithmetic theory remains subtler than its classical counterpart precisely because the Y-diamond rule does not turn positive seed data into automatic arithmeticity (Germain, 2024).