Independent Plaquette Model Overview

• Independent Plaquette Model is a family of constructions where plaquettes, rather than bonds or sites, serve as the primary degrees of freedom in statistical and quantum systems.
• The model features variants like the Bernoulli measure baseline, decoupled quantum clusters, and constrained plaquette spins, each highlighting distinct regimes of interaction.
• It provides insight into complex dynamics and phase transitions by contrasting ideal independence with modifications from topology, boundary conditions, and parity constraints.

Searching arXiv for recent and relevant papers on plaquette models and the independent plaquette viewpoint. The independent plaquette model is not a single universally standardized model, but a family of plaquette-based statistical-mechanical constructions in which the elementary energetic or probabilistic degrees of freedom are attached to plaquettes rather than bonds or sites. Across the literature, the expression is used in at least three closely related senses: as the $q=1$ Bernoulli baseline underlying the plaquette random-cluster model (PRCM), where $i$-plaquettes are open independently with probability $p$; as the decoupled-cluster limit $t'\to 0$ of plaquette Hubbard or plaquette Heisenberg systems, where local plaquettes become autonomous quantum units; and as a heuristic description of plaquette spin models, such as the square plaquette model (SPM), whose defect variables have simple plaquette weights but remain globally constrained (Duncan et al., 2024). In all cases, the central idea is that plaquettes provide the natural local objects, but complete independence is typically exact only in a limiting or baseline sense; the full models generally incorporate parity constraints, topological weights, inter-plaquette couplings, or subsystem symmetries that qualitatively alter equilibrium structure and dynamics (Chleboun et al., 2019).

1. Conceptual definition and main variants

In the probabilistic setting of higher-dimensional percolative systems, the independent plaquette model is the product Bernoulli measure in which each $i$-plaquette is open independently with probability $p$. The PRCM is defined on subcomplexes $P$ satisfying

$X^{i-1}\subset P\subset X^i,$

with probability measure

$$\mu_{X,p,q,i}(P)\propto p^{P}(1-p)^{X^{i}-P}\,q^{\,H^{i-1}(P;\mathbb{Z}_q)},$$

or equivalently

$$\hat{\mu}_{X,p,q,i}(P)\propto p^{P}(1-p)^{X^{i}-P}\,q^{\,H_{i-1}(P;\mathbb{Z}_q)}.$$

Here the independent plaquette model appears precisely as the limiting case $i$0, for which the topological factor is trivial and the plaquettes are independent (Duncan et al., 2024).

In classical spin systems, the phrase is often used more heuristically. In the SPM, the Hamiltonian is a sum of plaquette products,

$i$1

with plaquette variables

$i$2

The induced measure on plaquette configurations has weight proportional to $i$3, so thermodynamically the defect field resembles a low-density gas of plaquettes. However, the admissible plaquette configurations satisfy parity constraints, so the model is not fully independent (Chleboun et al., 2019, Chleboun et al., 2018).

In quantum lattice models, “independent plaquette” usually denotes a decoupled-cluster limit. In the plaquette Hubbard Hamiltonian, the independent plaquette limit is

$i$4

where the system breaks into disconnected $i$5 Hubbard clusters (Ying et al., 2014). In the orthogonal-plaquette Heisenberg model, exact product states built from local four-spin singlets realize an analogous independent-plaquette physics, because inter-plaquette couplings annihilate plaquette singlets in suitable parameter regimes (Boos et al., 2019).

A plausible implication is that the term functions more as a structural paradigm than as a unique model definition: it identifies regimes where the plaquette is the primitive object and interactions between plaquettes are absent, weak, or reduced to explicit constraints.

2. Probabilistic and topological formulation

The most explicit formalization of the independent plaquette baseline appears in the PRCM, which is presented as a higher-dimensional analogue of the Fortuin–Kasteleyn random-cluster model (Duncan et al., 2024). Its state space consists of all choices of $i$6-plaquettes, with all $i$7-cells always present. The independent plaquette model corresponds to the product Bernoulli law on these $i$8-plaquettes.

The PRCM modifies that baseline by inserting a topology-dependent cluster weight. The factor

$i$9

means $p$0 in the intended counting sense, so the probability of a configuration depends not only on how many plaquettes are open, but also on the cohomology or homology of the resulting subcomplex (Duncan et al., 2024). Plaquettes therefore cease to be independent: opening a plaquette changes the weight according to its effect on homological structure.

This formulation yields a direct and mathematically precise statement of what independence means in plaquette systems. At $p$1, the model reduces to independent plaquette percolation. At $p$2, the model becomes a topologically weighted extension. The paper emphasizes that the cohomological and homological formulations agree on the relevant subcomplexes by the universal coefficient theorem (Duncan et al., 2024).

Boundary conditions are correspondingly nontrivial. For a box $p$3, a boundary condition $p$4 determines

$p$5

and the finite-volume law is

$p$6

Free boundary conditions correspond to closing exterior plaquettes, while wired boundary conditions are expressed through Borel–Moore homology and identify boundary $p$7-cycles in the appropriate topological sense (Duncan et al., 2024). This shows that even the notion of “independence” is volume- and boundary-dependent once topological couplings are introduced.

3. Independence versus constraints in plaquette spin models

In plaquette spin models, the independent-plaquette picture is exact at the level of local energetic accounting but fails globally because the plaquette variables are not free. This is clearest in the SPM, where the plaquette variable $p$8 is called a defect, and the induced measure on plaquette configurations is

$p$9

The weight depends only on the number of defects $t'\to 0$0, subject to admissibility constraints (Chleboun et al., 2019).

For plus or periodic boundary conditions, the number of defects in each row and each column is even (Chleboun et al., 2019). Likewise, in the critical-scale analysis of the SPM, periodic parity constraints are written as

$t'\to 0$1

for all columns and rows (Chleboun et al., 2018). Thus the field of defects resembles an independent low-density gas only after conditioning on global parity restrictions.

This distinction matters for both equilibrium and dynamics. The dynamics are continuous-time single-spin Glauber/Metropolis dynamics,

$t'\to 0$2

and a single spin flip changes the four surrounding plaquette variables (Chleboun et al., 2019, Chleboun et al., 2018). The induced defect dynamics are therefore constrained particle dynamics rather than independent plaquette flips.

The independent-plaquette approximation remains useful at low temperature because defects are rare. However, the cited works stress that parity constraints and geometric structure are decisive, especially under periodic boundary conditions, where they generate a large degenerate ground-state manifold and effective reduced dynamics on that manifold (Chleboun et al., 2019).

A common misconception is that plaquette spin models are literally equivalent to independent defect gases. The literature summarized here does not support that stronger claim. The precise statement is instead that the energy is naturally expressed in plaquette variables, while admissibility and dynamics remain nontrivial.

4. Dynamical consequences: the square plaquette model

The SPM is a canonical example of how the independent-plaquette viewpoint clarifies thermodynamics but does not by itself determine kinetics. The model is studied on the critical length scale

$t'\to 0$3

described as the correlation length for products of spin variables in the infinite-volume Gibbs measure (Chleboun et al., 2019, Chleboun et al., 2018).

With all-plus boundary conditions, there is a unique ground state, the all-$t'\to 0$4 configuration. With periodic boundary conditions, there are

$t'\to 0$5

ground states, obtained by flipping arbitrary collections of rows and columns relative to the all-plus configuration (Chleboun et al., 2019, Chleboun et al., 2018). On the critical scale, ground states dominate the stationary measure: $t'\to 0$6 in the critical-scale mixing analysis, and

$t'\to 0$7

in the cutoff analysis as quoted in the supplied source block (Chleboun et al., 2018, Chleboun et al., 2019).

The dynamical timescales depend strongly on boundary conditions. For the SPM on the critical scale, one set of results gives

$t'\to 0$8

for all-plus boundary conditions, and

$t'\to 0$9

for periodic boundary conditions (Chleboun et al., 2018). A later refinement proves cutoff in the periodic case: $i$0 with

$i$1

for every $i$2 (Chleboun et al., 2019).

The mechanism is not independent defect relaxation. Under periodic boundary conditions, the trace chain on ground states is encoded by a bijection

$i$3

so the effective slow mode is a perturbation of random walk on a $i$4-dimensional hypercube (Chleboun et al., 2019). This demonstrates that constraints inherited from the spin-to-plaquette map dominate the low-temperature relaxation.

5. Independent plaquettes in quantum lattice models

In quantum many-body systems, the independent plaquette model usually denotes a decoupled-cluster limit or an exact plaquette-product regime. The plaquette Hubbard Hamiltonian is

$i$5

Its independent plaquette limit is

$i$6

where the lattice decomposes into disconnected $i$7 Hubbard clusters (Ying et al., 2014).

This limit serves as one endpoint of a nonmonotonic pairing problem. For $i$8, determinant quantum Monte Carlo finds the strongest $i$9-wave pairing signal near

$p$0

rather than at $p$1 or $p$2, and the optimal $p$3 increases with $p$4 (Ying et al., 2014). The decoupled plaquette limit therefore captures strong local singlet tendencies but lacks inter-plaquette coherence.

A related independent-plaquette principle appears in the orthogonal-plaquette Heisenberg model, whose Hamiltonian is

$p$5

The model admits two exact plaquette-singlet product ground states,

$p$6

and

$p$7

with exact energies per spin

$p$8

The exactness mechanism is that inter-plaquette couplings cancel when neighboring units are plaquette singlets (Boos et al., 2019).

This suggests a useful distinction. In probabilistic models, independence is usually a product measure on plaquettes. In correlated quantum models, independence more often means exact factorization into local plaquette singlets or decoupled plaquette clusters.

6. Degeneracy, subsystem symmetries, and layered order

In the three-dimensional plaquette Ising model, the plaquette formulation produces a different type of independent-plaquette intuition: not product-measure independence, but an enlarged low-energy manifold generated by plane flips. The Hamiltonian is

$p$9

with spins $P$0 on the vertices of a cubic lattice (Johnston et al., 2016, Johnston et al., 2015).

A key feature is that flipping all spins on an entire lattice plane costs no energy. For the pure plaquette model $P$1, any planes may be flipped, leading on an $P$2 lattice to ground-state degeneracy

$P$3

This degeneracy persists into the low-temperature phase (Johnston et al., 2016, Johnston et al., 2015). Because whole planes can flip independently, the ordinary magnetization

$P$4

vanishes as an order parameter (Johnston et al., 2016).

The anisotropic “fuki-nuke” limit clarifies the resulting order. Setting $P$5 yields a model that can be rewritten using bond variables

$P$6

reducing essentially to a stack of decoupled 2D Ising models, up to constraints expected to disappear in the thermodynamic limit (Johnston et al., 2015). This motivates layered order parameters such as

$P$7

which capture planar ordering invisible to the conventional magnetization (Johnston et al., 2016, Johnston et al., 2015).

The same degeneracy modifies first-order finite-size scaling. Because $P$8, the leading finite-size shift changes from the standard $P$9 form to $X^{i-1}\subset P\subset X^i,$0, for example

$X^{i-1}\subset P\subset X^i,$1

(Johnston et al., 2016). Numerical analyses reported for the isotropic model are consistent with this modified scaling, including

$X^{i-1}\subset P\subset X^i,$2

(Johnston et al., 2015).

A plausible implication is that plaquette-based models often replace conventional symmetry breaking by subsystem or layered structures, so “independence” manifests through decoupled planes or local sectors rather than through factorized Gibbs measures.

7. Interpretive role, misconceptions, and related directions

The literature distinguishes carefully between an independent plaquette baseline and the actual interacting models built from it. In the PRCM, the baseline is exact only at $X^{i-1}\subset P\subset X^i,$3; for $X^{i-1}\subset P\subset X^i,$4, topology-dependent cluster weights produce nontrivial dependence and duality between an $X^{i-1}\subset P\subset X^i,$5-dimensional and a $X^{i-1}\subset P\subset X^i,$6-dimensional plaquette model with parameter

$X^{i-1}\subset P\subset X^i,$7

(Duncan et al., 2024). In the SPM, the low-density defect gas picture is useful but incomplete because parity constraints and the spin-to-plaquette map determine admissibility and drive the dynamics (Chleboun et al., 2019, Chleboun et al., 2018).

In glassy plaquette Ising models, a further misconception is to identify slow relaxation solely with degeneracy. The three-dimensional plaquette Ising model exhibits strong metastability, slow coarsening, aging, and cooling-rate effects, and Monte Carlo simulations show the autocorrelation

$X^{i-1}\subset P\subset X^i,$8

is well fit in the supercooled regime by

$X^{i-1}\subset P\subset X^i,$9

with $$\mu_{X,p,q,i}(P)\propto p^{P}(1-p)^{X^{i}-P}\,q^{\,H^{i-1}(P;\mathbb{Z}_q)},$$0 (Lipowski, 2012). The same work argues that a related gonihedric model with a strongly degenerate ground state but lacking glassy features does not exhibit such a decay, indicating that degeneracy alone is not sufficient (Lipowski, 2012).

The independent plaquette idea also enters models with alternating plaquette structure rather than literal decoupling. In the plaquette orbital model, square-lattice faces are partitioned into $$\mu_{X,p,q,i}(P)\propto p^{P}(1-p)^{X^{i}-P}\,q^{\,H^{i-1}(P;\mathbb{Z}_q)},$$1- and $$\mu_{X,p,q,i}(P)\propto p^{P}(1-p)^{X^{i}-P}\,q^{\,H^{i-1}(P;\mathbb{Z}_q)},$$2-plaquettes in a checkerboard pattern, and local plaquette-flip symmetries imply that magnetic order is forbidden while orientational long-range order survives (Biskup et al., 2010). The paper argues that the observed checkerboard plaquette-energy pattern is an artifact of the observable choice and that the true order is orientational, not magnetic (Biskup et al., 2010).

Taken together, these results define the independent plaquette model not as a single canonical Hamiltonian, but as a recurring reference construction. It denotes the regime in which plaquettes are the primitive local objects and are independent, decoupled, or exactly factorized; the main physics of interest then lies in understanding how topology, constraints, inter-plaquette couplings, boundary conditions, or subsystem symmetries deform that baseline into the full correlated model (Duncan et al., 2024).