Star–Star Relation in Lattice Models

• Star–star relation is a local integrability condition that equates two distinct four-edge star configurations using renormalized IRF weights.
• It is realized in various lattice models—such as the checkerboard hyperbolic and two-component elliptic lens models—with weights constructed from specialized gamma functions.
• The relation bridges integrable lattice systems and gauge theory dualities, underpinning quasi-classical expansions and the formulation of multicomponent discrete 5-point equations.

Searching arXiv for recent and foundational papers on the star–star relation in integrable lattice models. The star–star relation is a local functional equation for Boltzmann weights in integrable lattice models. In checkerboard edge-interaction and interaction-round-a-face (IRF) settings, it relates two different four-edge star configurations, is a sufficient integrability condition, and, after a suitable renormalization of IRF weights, yields a Yang–Baxter equation of IRF type (Kels, 2024). In modern treatments it appears for multicomponent real spins, for mixed continuous/discrete spins built from lens elliptic or lens hyperbolic gamma functions, and within the Gauge/YBE correspondence, where it is identified with equalities of supersymmetric partition functions or indices under duality transformations (Kels, 2015, Catak et al., 27 Aug 2025).

1. Formal definition and local structure

In the general checkerboard setting, the star–star relation compares a black-centered star $V^{(B)}$ and a white-centered star $V^{(W)}$, each obtained by integrating over one internal spin. Its explicit form is

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$

This equation relates the two star configurations up to extra edge factors between same-color vertices, and in the formulation of the model it is a sufficient integrability condition (Kels, 2024).

A renormalized IRF weight $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$ can be defined from $V^{(B)}$ or equivalently from $V^{(W)}$. By virtue of the star–star relation, this renormalized weight satisfies an IRF Yang–Baxter equation. In this sense, the star–star relation is an analogue of, and is equivalent after reinterpretation of weights to, a Yang–Baxter equation (Kels, 2024).

In contrast to the star–triangle relation, which involves three edges and three spins, the star–star relation involves two four-edge stars and six boundary spins plus two internal spins. In IRF language it can be rephrased as equality of partition functions of two small local configurations. That distinction is structurally important because some constructions start from star–star identities and only later derive star–triangle relations or reduced symmetries from them (Kels, 2024, Catak et al., 27 Aug 2025).

2. Lattice realizations and spin content

A representative multicomponent realization is an edge-interaction model on a checkerboard square lattice. Spins live on vertices and interactions sit on edges. Each spin at vertex $i$ is an $n$-component real vector

$$\mathbf{x}_i=\bigl((x_i)_1,\ldots,(x_i)_n\bigr)\in\mathbb{R}^n, \qquad \sum_{a=1}^n (x_i)_a=0,$$

so there are $n-1$ independent components. The case $V^{(W)}$0 is scalar. The lattice carries horizontal and vertical rapidity lines with parameters $V^{(W)}$1 and $V^{(W)}$2, and the four edge types carry weights $V^{(W)}$3, $V^{(W)}$4, $V^{(W)}$5, and $V^{(W)}$6 in the pattern fixed by the checkerboard construction (Kels, 2024).

From these edge weights one constructs the black-centered and white-centered IRF weights by integrating over an internal spin with a single-spin weight $V^{(W)}$7. The star–star relation then equates the two local configurations. This is the canonical IRF-style formulation of the relation in the multicomponent hyperbolic setting (Kels, 2024).

A distinct but closely related realization uses two-component spins

$V^{(W)}$8

or, in the main elliptic lens model,

$V^{(W)}$9

Here the weights are built from the lens elliptic gamma function, and the model belongs to the mixed discrete/continuous class emphasized in the Gauge/YBE correspondence (Kels, 2015).

Setting	Spin variables	Functional input
Checkerboard hyperbolic model	$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$0	Hyperbolic gamma function $$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$1
Two-component lens model	$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$2 with continuous and discrete parts	Lens elliptic gamma function
Reduced two-spin symmetry	$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$3	Lens hyperbolic gamma, basic hypergeometric, or Euler gamma functions

These realizations show that the term “star–star relation” does not denote a single scalar identity. Rather, it denotes a class of local relations whose common role is to encode integrability for IRF or multi-spin models, while the spin space and special-function input vary from elliptic to hyperbolic, trigonometric, and rational regimes (Kels, 2024, Kels, 2015, Catak et al., 27 Aug 2025).

3. Explicit special-function solutions

In the multicomponent hyperbolic solution, the crossing parameter is

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$4

and the spins and rapidities satisfy

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$5

The single-site weight is

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$6

and the two-spin weight is

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$7

These weights satisfy both inversion,

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$8

and crossing symmetry,

$$W_{q'-q}(\mathbf{x}_d,\mathbf{x}_c)\,W_{q-q'}(\mathbf{x}_a,\mathbf{x}_b)\, V^{(B)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d) = W_{p-p'}(\mathbf{x}_b,\mathbf{x}_d)\,W_{p'-p}(\mathbf{x}_c,\mathbf{x}_a)\, V^{(W)}(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d).$$9

and the corresponding star–star relation is written as an identity between the two IRF-type integrals $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$0 and $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$1 (Kels, 2024).

In the two-component elliptic lens model, the weights are built from the lens elliptic gamma function $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$2. The edge Boltzmann weight has the explicit form

$$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$3

with a corresponding single-spin weight $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$4. In this setting the weights satisfy spin reflection symmetry, are normalized so that the factor $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$5 in the star–triangle relation equals $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$6, and are real and positive for $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$7 and $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$8 (Kels, 2015).

The elliptic construction is tied to Yamazaki’s star–star relation, whose Boltzmann weights are built from the lens elliptic gamma function and arise in the Gauge/YBE correspondence from Seiberg duality of $$V(\mathbf{x}_a,\mathbf{x}_b,\mathbf{x}_c,\mathbf{x}_d)$$9-d $V^{(B)}$0 quiver gauge theories. Kels’ construction starts from Yamazaki’s two-component star–star solution and extracts explicit edge weights suitable for an Ising-type lattice model (Kels, 2015).

The same family of constructions admits degenerations. The elliptic lens star–triangle relation contains the Bazhanov–Sergeev master solution as a special case; the hyperbolic multicomponent star–star relation is equivalent to hyperbolic hypergeometric integral identities for the $V^{(B)}$1 root system; and later work derives reduced flipping symmetries with lens hyperbolic gamma, basic hypergeometric, and Euler gamma functions (Kels, 2015, Kels, 2024, Catak et al., 27 Aug 2025).

4. Relation to the star–triangle relation and to flipping

For Ising-type models with discrete and continuous spins, the star–triangle relation takes the form

$V^{(B)}$2

with

$V^{(B)}$3

In Kels’ two-component lens model, the normalization is chosen so that $V^{(B)}$4, and the resulting star–triangle relation implies the corresponding two-component spin star–star relation of Yamazaki. The paper explicitly emphasizes that the former relation implies the latter, but the reverse is not true (Kels, 2015).

This implication is central to the role of the star–star relation in that work. Yamazaki’s construction produces a star–star relation in IRF language from gauge-theory duality, while Kels proves a star–triangle relation for the corresponding edge weights. The star–star relation then appears as a corollary of a stronger local identity (Kels, 2015).

A later reduction produces the flipping relation. In the lens hyperbolic setting, the flipping relation is

$V^{(B)}$5

It exchanges the edge interactions of two outer spins with a centrally sited spin, and a certain limit of the lens hyperbolic gamma star–star relation yields this symmetry transformation. The same work obtains further solutions of the flipping relation in terms of the hyperbolic gamma, basic hypergeometric, and Euler gamma functions (Catak et al., 27 Aug 2025).

This hierarchy is structurally informative. The star–star relation can act as the primary IRF identity, the star–triangle relation can be extracted or proved in special cases, and the flipping relation can arise as a reduced symmetry obtained by sending selected fugacities or spins to infinity in the corresponding hyperbolic integral identity. A plausible implication is that these relations are best viewed as different local manifestations of the same integrable architecture rather than as isolated identities (Kels, 2015, Catak et al., 27 Aug 2025).

5. Quasi-classical expansion and multicomponent 5-point equations

The quasi-classical expansion of the hyperbolic multicomponent star–star relation is organized by

$V^{(B)}$6

together with the scaling

$V^{(B)}$7

Using the asymptotics of $V^{(B)}$8, one obtains

$V^{(B)}$9

with explicit classical Lagrangians $V^{(W)}$0 and $V^{(W)}$1 (Kels, 2024).

Substituting these expansions into the star–star relation and applying a saddle-point analysis produces $V^{(W)}$2 component equations for the internal spin. In multiplicative variables, the central rational function is

$V^{(W)}$3

and the quasi-classical equations are

$V^{(W)}$4

These are $V^{(W)}$5-component 5-point difference equations on a plus-shaped stencil (Kels, 2024).

For $V^{(W)}$6, the system reduces to scalar equations. The paper identifies the scalar reduction of the hyperbolic system with the multiplicative four-leg form of the 5-point equation $V^{(W)}$7, and the rational degeneration with $V^{(W)}$8. For general $V^{(W)}$9, the equations are $i$0-component extensions of these scalar 5-point equations (Kels, 2024).

At the level of multidimensional consistency, the quasi-classical limit of the IRF Yang–Baxter equation yields a system of $i$1 multicomponent 5-point equations on a face-centered cubic cell, providing consistency-around-a-face-centered-cube (CAFCC). Numerical verification of CAFCC is reported for $i$2. This establishes a direct bridge from the star–star relation to multicomponent discrete integrable systems with a variational structure (Kels, 2024).

6. Gauge/YBE correspondence, scope, and terminology

Within the Gauge/YBE correspondence, the star–star relation is identified with equalities of supersymmetric partition functions or indices. Yamazaki’s star–star relation arises from the Gauge/YBE correspondence by translating Seiberg duality of $i$3-d $i$4 supersymmetric indices into an IRF-type functional identity. Kels’ star–triangle relation implies Seiberg duality for the $i$5-d $i$6 $i$7 index of the $i$8 quiver gauge theory, and therefore also the corresponding two-component star–star relation (Kels, 2015).

In the lens hyperbolic setting, the same theme reappears for $i$9-d $n$0 supersymmetric gauge theories on $n$1. The star–star relation is realized as an identity between partition functions built from lens hyperbolic gamma functions, and the flipping relation appears as a reduced star–star relation obtained by a limit that decouples part of the original configuration (Catak et al., 27 Aug 2025).

The mathematical significance of these identities is standard across the cited works. They guarantee commutativity of transfer matrices and hence integrability; in IRF language they give a Yang–Baxter equation for renormalized face weights; and in gauge-theoretic language they encode duality transformations of localized partition functions (Kels, 2024, Kels, 2015).

The phrase “star–star relation” also has unrelated uses outside integrable lattice models. In astrophysical literature, the phrase may refer broadly to empirical relations among stellar variables such as the mass–luminosity, mass–radius, and mass–effective temperature relations, or to star-formation scaling relations. Those usages concern statistical relations among stellar or galactic observables and are conceptually distinct from the Yang–Baxter-type meaning discussed here (Eker et al., 2024, Schruba, 2012).

Taken in its integrable-model sense, the star–star relation is therefore best understood as a local IRF identity with several tightly connected realizations: elliptic and hyperbolic special-function solutions, reductions to star–triangle and flipping relations, quasi-classical limits yielding multicomponent 5-point equations, and gauge-theoretic interpretations through supersymmetric dualities (Kels, 2024, Kels, 2015, Catak et al., 27 Aug 2025).