WorksMathematics
Boxes with cyclic tiling groups have one layer or size 2 × 2 × 4
Abstract
Kuperberg asked how a box's dimensions determine its tiling group. We prove Conlon's conjecture that this group is cyclic exactly when one side has length one or the box has dimensions in some order. For every box, we also show that the minimum number of generators of its tiling group equals the largest multiplicity of a cyclotomic factor of prime-power index in MacMahon's volume polynomial. This polynomial counts stacks by the number of cubes they contain. It also determines the Smith factors over of the box's weighted Carlitz matrix, which presents the tiling group at . Each cyclotomic factor of multiplicity occurs in exactly the last Smith factors, once in each.
Which boxes have a cyclic tiling group
Stack unit cubes in a corner of an box so that column heights weakly decrease away from that corner in both horizontal directions. This boxed plane partition projects to a hexagon tiled by diamonds called lozenges (Figure 1). The matrix of possible tile placements presents a finite abelian group , the tiling group, with one generator for each row and one relation for each column (Figure 2). Its order is , the number of possible stacks in that box [14 p. 265]. In 1998 Kuperberg asked which group of that order occurs for each box [8 Question 2].

Figure 1. One of the 105 stacks in the box. Its four columns have heights 4, 2, 1, 0. Viewed along a body diagonal, each visible unit square, whether the face of a cube or part of the box’s floor or walls, projects to a lozenge. Together these lozenges tile the hexagon on the right, with the floor and walls shown in pale colors. The dotted segment in the bottom lozenge is its short diagonal. It determines that tile uniquely because the two unit equilateral triangles with this common side form the lozenge.
Figure 2. The marked diagonals of each tiling select one 1 in every row and column of . Triangle A shares a lozenge with in the empty box and with in the full box. In column A these two lozenges give a circled 1 in row and a boxed 1 in row . Adding those row labels gives the group relation .
Theorem 1 (Conlon’s conjecture on cyclic tiling groups). For positive integers , the group is cyclic if and only if some side has length one or is a permutation of . The exceptional group is isomorphic to .
Theorem 1 tells us when one generator suffices. How many generators does an arbitrary box need? Give each stack of cubes weight and sum over all stacks to obtain MacMahon’s volume polynomial [11 §429]. Write for the multiplicity in of , the cyclotomic polynomial whose roots are the primitive th roots of unity.
Theorem 2 (Minimum number of generators). The tiling group needs exactly generators, where runs over the primes and over the positive integers.
Kuperberg observed that the tiling group of a one-layer box is cyclic [8 Question 2]. In 2002 he proposed two conjectures that together would determine the group for every box [9 §5.1, Conjectures 11–12]. Conlon found the exceptional box by computer and conjectured that it was the only exceptional case [3 §7]. MacMahon’s count of the stacks [11 §429] gives the group order , so by Cauchy’s theorem there are elements of orders , , and . Since these elements commute and their orders are pairwise coprime, their product has order and generates the group. § 6 rules out the remaining boxes by a gcd calculation when the shortest side is two and by a bound on element orders when it is larger. § 2 illustrates Theorem 2 on two boxes with twenty stacks, and § 7 proves it.
From the picture to the group
The three orientations of a lozenge correspond to squares parallel to the floor or to either wall. Projection gives a bijection between stacks in an box and lozenge tilings of the equiangular hexagon with successive side lengths , , , , , [9 §5.1, Fig. 10].
Marking the short diagonal of every lozenge records the whole tiling. To list the possible marks, divide the hexagon into unit triangles. In the orientation of Figure 1 and Figure 2, the right-pointing triangles () index the rows of a matrix and the left-pointing ones () its columns. Put 1 for a shared side, the short diagonal of a possible tile, and 0 otherwise. A choice gives a tiling exactly when its tiles cover each triangle once, selecting one 1 per row and column [14 p. 261]. The Kasteleyn–Percus formula counts these tilings by [14 p. 265].
The generators of are indexed by the rows of . For each column, set the sum of the generators at its 1’s equal to zero. In the one-cube box (Figure 2), column gives , and then column gives . Substituting in column leaves one generator with the single relation .
With triangles of each orientation, this construction gives the quotient , the integer cokernel of . Its order is the lattice index . Kuperberg also asked how its elements could be interpreted in terms of the tilings [14 p. 265, Problem 5].
Does the stack count determine cyclicity?
Both boxes in Figure 3 have twenty stacks, and only the flat box’s tiling group is cyclic. Each colored line traces the boundary of a layer halfway up its cubes, continuing along the box walls where needed. Adding a unit cube interchanges two consecutive steps in its layer’s path and leaves its endpoints fixed. The shortest paths between opposite floor corners give all the flat stacks. In the pictured box, three unit cubes form an L and a fourth sits above its corner. Figure 4 shows all twenty stacks in this box.
Figure 3. Both boxes have twenty stacks. needs one generator and needs two. The colored paths trace the layers on the dotted grid. The lower right records the coefficient parities in , so gives (even, odd). Adding changes only the parity of , and adding only that of .
Figure 4. The twenty stacks in the box, arranged by number of cubes. The darker faces belong to the cubes and the pale faces belong to the box. Setting counts all twenty stacks.
Kuperberg’s path matrix presents the same group as the triangle matrix [9 Corollary 7 and Theorem 14]. Its rows index path starts (generators) and its columns index path ends (relations). Each entry counts paths between its endpoints by rising or falling steps to the right. The flat box gives . In the two-layer box, two of four steps rise for a level finish (six choices), and one or three rise for a change of height (four choices).
The upper path start gives and the lower gives . In , both relations preserve the parities of because their coefficients are even. All four pairs occur at . Multiples of any give only the parities of or , so two generators are necessary.
What the volume polynomial remembers
A cyclic group contains an element of order . To rule this out for every box with shortest side at least three, we bound the orders of its elements using the volume polynomial. Figure 4 sorts the twenty stacks in the box by volume. A rule that always adds or removes one unit cube sends all twelve even-volume stacks into the eight odd-volume stacks. Such a rule cannot realize addition by a group element, which is a bijection.
The four stacks in the middle column contribute . Reading across the figure gives the volume polynomial
where and . The factor has prime-power index and occurs twice, so by Theorem 2 the group needs exactly two generators. For the flat box, MacMahon’s product gives . Each factor of prime-power index occurs once, so one generator suffices. Both polynomials give at .
The map respects both relations for the box in Figure 3 and takes all twenty values. Since the group has order twenty, this gives . To see these summand orders in the volume polynomial, number the four steps of each path in Figure 3 from to . Give a rise weight and a fall at position weight , and multiply the weights along the path. Choosing in records a fall at , so the coefficient of counts paths with falls by weight. From the upper start to the lower endpoint there are three falls, giving in the upper-right entry. At every path has weight one.
The determinant of is times a power of . Allow rational coefficients and powers of of either sign. Invertible row and column operations then give a diagonal matrix with each entry dividing the next. These entries are the Smith factors, determined up to nonzero rational constants and powers of .
The first Smith factor divides the second, so its square divides . Since and are distinct irreducibles, the only possibilities are and . At , where , the first has rank one and the second has rank zero. The row and column operations remain invertible at , so we can read the rank from .
At this value the product defining the entries becomes . All three middle coefficients vanish, so . This selects , giving the summand orders and at .
The rule for every box
For the general box, fix the lengths , , , put , and let be the coefficient of in , with outside . The weighted Carlitz matrix and MacMahon’s volume formula [11 §429] are
Kuperberg relates , the dual Jacobi–Trudi matrix of the rectangle , to the weighted tiling matrix by invertible row and column operations and adjoining identity blocks. He asks about a diagonal form with squarefree entries over the integral Laurent ring [9 Theorem 14 and Conjectures 11–12]. For ordinary boxes we determine the factors over the rational Laurent ring. If an irreducible factor occurs times in the volume polynomial, it goes in different diagonal entries. Since each entry divides the next, these must be the last .
Theorem 3. Let denote the -th cyclotomic polynomial and put . In increasing divisibility order, the Smith factors of over , up to units, are
For the box, occurs twice and occurs once. For arbitrary boxes, the operations used over the rational Laurent ring can introduce integer denominators. To draw conclusions about the tiling group at , § 5 therefore proves divisibility for the determinants of square submatrices before setting .
The rank test at suggests how to prove the weighted theorem. At a primitive -th root of unity the determinant vanishes to order . Each diagonal entry that vanishes uses at least one of those factors and lowers the rank by one. We will find independent vectors sent to zero, forcing all copies into different entries.
The two intervals of exponents
Write the coordinates of a vector as the coefficients of a polynomial of degree . Put . Multiplying by and taking the coefficients of gives up to row and column signs. Indeed, coefficient in this block applied to is . The coefficient map sends a vector to zero precisely when these middle coefficients vanish. Its product can then have nonzero terms only in the two outside intervals. Transposition and the signs preserve rank. Thus at any nonzero specialization of to a field, the nullity of is the dimension of the polynomials supported on
that are divisible by . Division recovers , since has degree and . For the box the allowed exponents are . At we have , so both and have the prescribed zeros and allowed support. They are independent. At a primitive fifth root, divides , which again has allowed support. These give the two missing ranks at a fourth root and the one missing rank at a fifth root, exactly the multiplicities we saw in the volume polynomial. Figure 5 shows how to find the multiples before doing any division.

Figure 5. A line joins two allowed powers whose exponents agree modulo the order of the root. Their differences on the right are the supported products . At the pale positions the coefficients of must vanish.
The same support gives a short determinant calculation. Put . Use as columns the monomials , in increasing order, followed by , in the monomial basis through degree . Expanding along the monomial columns gives . Alternatively, use as coordinate basis the monomials from followed by the same multiples of . The change of basis has determinant . In these coordinates the only determinant left is that of the remainders of , , modulo . Over , evaluation at , , computes this last determinant as a quotient of Vandermonde determinants, one on and the other on . Pairs within each interval cancel. The cross pairs give , and the powers of give
Indeed, the cross differences in the numerator and denominator have lengths and for , . Shifting the last exponents of by contributes the monomial to the Vandermonde quotient.
Fix , write , , and let be the number of elements of congruent to mod . Every difference of distinct exponents contributes one factor precisely when its two exponents agree modulo . Among the exponents in , residue classes have size and the others have size . Hence
The second identity uses and the polynomial convention , also at .
Finding all the missing ranks
Let be a primitive -th root of unity. At one full period of factors in gives . Thus
This gives two ways to construct a multiple of with support in . We can impose full periods of zeros, or impose full periods and then the remaining zeros.
The two rows of Figure 5 show why we use both constructions. At , one full period gives the two multiples and . A second full period would require the factor , whose degree exceeds the largest allowed exponent . At , counting four root conditions on four coefficients gives only the lower bound zero. Completing the period gives the supported multiple .
Lemma 4. With , the nullity at is at least
Proof. Write a supported polynomial uniquely as . The coefficients of have degrees of freedom. Requiring imposes at most linear conditions, namely zero remainder modulo . The resulting space has dimension at least , and every in it is divisible by . Choosing gives the first bound. Choosing , divide by and require zero remainder modulo . This imposes at most further conditions and gives the second bound.
The shape of the box now enters. Each of the two intervals in (1) contributes either the floor or the ceiling of its length divided by to . Write , , with .
If , then and every is , , or . In (3) only the ’s contribute. Their number is exactly the first lower bound in Lemma 4.
If , then and every is , , or . Only the ’s contribute to (3). Their number is , since . This is the second lower bound. These arguments also cover and .
We have found at least dimensions of kernel. In Smith form the nullity at counts the invariant factors divisible by . Their valuations sum to by (2), so there are exactly of them, each of valuation one. Divisibility puts them in the last positions. All nonunit irreducibles of the determinant are cyclotomic, which proves Theorem 3.
There are exactly nonunit rational Laurent factors. For , at a primitive -th root the product defining the elementary functions has only its constant and top coefficients, and every matrix entry is intermediate. Thus the entire matrix vanishes. Permuting the box sides preserves , so the number has the same value in every orientation. This also gives the count proposed in [3 §9] for an integral Laurent Smith form if it exists, since primitivity of the determinant precludes additional nonunit constant factors.
Consequences for the integer matrix
Let , the quotient of by the image of the integer Carltiz matrix. Its order is . For put
Over the product of the first Smith factors divides every -rowed minor. That product is . Its coprimality to lets us remove powers of over . Division by the monic polynomial then proves divisibility of the actual minors in . This makes specialization legitimate even if rational row operations would introduce denominators.
Corollary 5. Every -rowed minor of is divisible by . In particular, the exponent of divides
which divides .
Proof. The exponent of , the largest order of one of its elements, is its last integer Smith factor. This is divided by the gcd of the cofactors, with gcd 1 in size one. For the box the diagonal cofactors are 6 and the off-diagonal cofactors are . Their gcd is 2, so the largest element order is . In general divides the cofactor gcd. Dividing by gives the stated product. All indices occurring are at most . For , if is a positive power of a prime , and is 1 otherwise. This gives both the lcm bound and .
Proof of the cyclicity classification
Proof of Theorem 1. Order the sides . Permuting them gives stably equivalent presentations of the same tiling group [14 Problem 5]. If , the Carlitz matrix has a single entry and the group is cyclic.
If , the group is cyclic exactly when the entries of the two-by-two Carlitz matrix have gcd 1, since their gcd is its first Smith factor. Put . The diagonal entries are , and the other two are and . Their gcd is at least , because multiplication by makes all three entries integer multiples of . The lower bound increases with either or for and equals at . For it is whenever . Finally at , the gcds are respectively 2, 5, 1. Only the last box gives a cyclic group. Its matrix has determinant 105. The map sends both columns to zero and the class of to 1, giving an explicit isomorphism with .
Suppose now that . A finite abelian group is cyclic exactly when its exponent equals its order. We show that the order outgrows an upper bound for the exponent.
Write and . Each active power divides a numerator index in the symmetric version of MacMahon’s formula having height . These indices range from to . The number of active levels is at most the largest active , so Corollary 5 gives
For the second divisibility, every power of a prime beyond that divides some contributes to the valuation of the binomial, and the smaller powers are covered by . In the product for , the cells of the first row and last column together give one factor for every . Thus
It suffices that , and increases with .
For it suffices to compare with the upper bound for . Start at , increase along , and then increase at fixed . Each step adds one to , multiplying the comparison bound by four.
To prove by induction, observe that divides , and divides . Any prime power missing from the smaller lcm gives a factor in the binomial, and at most one power per prime is missing. Both binomials are at most , using the two equal central coefficients in the odd row. For , multiplying its two rows gives
The last inequality reduces to . Since , we have . Each new row up to increases by a factor of at least . Therefore .
For the first pair of columns below covers . The second covers every , by monotonicity of and .
| 4 | 60 | 420 | ||
| 5 | 924 | 420 | 13068 | 2520 |
| 6 | 840 | 27720 |
Table 1.
For and , . For the respective group orders are 980, 4116, 14112, while is 840, 2520, 2520. This proves noncyclicity for every box with minimum side at least three and completes the classification.
Another prediction in Conlon’s §7 concerns the number of generators for an box. It requires generators exactly when is a power of two, including . Indeed this happens when some prime divides every entry. The entries comprise all interior binomial coefficients of row . If , , then in . All interior coefficients vanish when , while for the coefficient at is . Hence must be a prime power, and the prime must be 2.
The ordinary group at a prime
The four coefficient-parity pairs in Figure 3 form for the box. Both relations vanish modulo 2, leaving the images of as a basis over . For any box and prime , reducing the generators and relations modulo gives , with one dimension for each nontrivial cyclic summand of the -primary subgroup .
Call active if , and let count the active indices. Corollary 5 gives and . Since is square, its nullity modulo equals . We compute it using Han’s theorem on relations between three powers of linear forms.
Theorem 6. For every the dimension of over is , with empty maximum zero. If , list the positive multiplicities and pad the list with zeros to obtain two numbers . Then
This gives the complete -primary group for every , and also for smaller primes having at most two active indices.
Deduction of Theorem 2. Write as a direct sum of cyclic groups of orders . One generator from each summand suffices. Any generating set of maps to a spanning set of . A prime dividing divides every , giving . The minimum for is therefore by Theorem 6.
To prove Theorem 6, we first express in terms of the side lengths modulo . Let be the least nonnegative residues of mod and . For ,
Indeed, with , the factorial form of the product in (2) gives
The identity permits reduction of each side length modulo . Since , the expression becomes
It vanishes for and . In the remaining range order and put . The expression is . As crosses , this takes the successive forms , giving (4).
For example, take a cubical box of side with . In (4) the only active powers of are and , with multiplicities and . At the intermediate powers all three residues are and their sum is less than . At higher powers even the original three sides sum to less than . Theorem 6 thus gives
In particular the box has four summands of order and one of order .
Proof of Theorem 6. Put , , , , and work over the algebraic closure of , which preserves the nullity. In , multiply by from degree to degree . The source monomials have -exponents and the target monomials have -exponents . Thus the matrix is . This use of multiplication in characteristic is related to the weak Lefschetz interpretation of prime divisibility of the box count in Li and Zanello [10 Theorem 4.2].
Let be the degrees of a homogeneous free basis for the relations between in . The relation module is free of rank two by the graded syzygy theorem in two variables. Its degrees satisfy because the Hilbert-series numerator of the finite-dimensional quotient has a double zero at . The third coefficient of a relation of degree runs through precisely the kernel of our map. That coefficient determines the whole relation, since the least degree of a relation on alone is . It follows that
Han’s theorem computes this degree gap by a distance. We use the statement in Brenner and Kaid [1 Theorem 2.3], which attributes the formula to Han [4] and also gives Monsky’s proof reference [13 Corollary 23]. Put and let be the triples of integers with odd sum. Write for the sum-of-absolute-values distance to a set. The coordinates of satisfy the strict triangle inequalities. If is the least integer for which , the gap is
It is zero when there is no such . Coordinate permutations preserve the formula. The sum of the coordinates of is even, so only can qualify. For any we claim that
Reducing the sides modulo changes by an integer triple with even sum. We can therefore use . If an odd center lies at distance , each of must be within of an odd multiple of . Indeed each is the sum of two coordinates of minus the third. In the required odd multiple can only be . The three equations on force . To compute the distance to this center, take the largest absolute signed coordinate sum, allowing all signs. Up to a common sign the four sums are , , , . Taking the positive part of minus their maximum absolute value gives twice the minimum in (4) when positive, and zero otherwise.
Han’s least therefore chooses the largest active power . It also gives the largest among the active powers. In fact, if are both active and , take their closest odd centers . The triangle inequality gives . This integer has parity , so it is at most . Hence
This proves the rank formula, also when no level is active. If at most two levels occur, the nontrivial cyclic -power factors have orders or , and the base- logarithm of the group order is . Thus exactly of those factors have order .
The first determinantal ideal over the integers
For the entries themselves one can strengthen the divisibility in Corollary 5 to an equality of ideals before specializing. The reason is that their indices run through a consecutive interval. Write for the Gaussian binomial coefficient. Up to a power of , for . Ekedahl’s full-interior-row result, published by Mirmohades [12 Theorem 3.1], gives the ideal when the indices are , . Hazi’s proof of Theorem 3.8 treats initial intervals in the two-colored setting [5]. Joris, Oestreicher and Steinig studied the gcd of consecutive ordinary binomial coefficients [7]. Here we use any nonempty consecutive interval in the polynomial ring.
Lemma 7. The for generate their monic polynomial gcd as an ideal in .
Proof. Let be their monic gcd in . The factorial formula for Gaussian binomials gives
Here , and these are all the irreducible factors. The quotients therefore belong to and have constant term 1. Suppose their ideal were proper. A rational polynomial Bézout identity gives a nonzero integer in that ideal after denominators are cleared, so a maximal ideal containing it has residue characteristic . Its residue field is a finite extension of . The constant terms force the image of to be nonzero. Denote its multiplicative order by . The cyclotomics vanishing there have indices , , with omitted from consideration. This follows from in characteristic and the distinct roots of when . For each such , call an index good if . If goodness is constant on for every , any attains all the minimum valuations. Otherwise take the largest on this chain for which there are both good and bad indices. Such exists because all indices are good for . The good indices modulo occur in the consecutive blocks . Since contains both kinds of index, it contains a good endpoint of one of these blocks. That endpoint is congruent to either or modulo every divisor of , so it is good for every smaller on the chain. At every larger level the choice of index is immaterial. Thus a single attains the minimum at every vanishing cyclotomic. Its quotient by is nonzero in the residue field, a contradiction.
For take in the integers. Powers of are units in , so the entries generate over the integral Laurent ring. In particular the smallest ordinary Smith factor always equals . The two-by-two weighted presentation also has a Smith form in each local ring at a maximal ideal, since after division by one entry is a unit and serves as a pivot.
Relation to earlier work. Kuperberg’s integral Laurent problem also asks that the diagonal form exist over . Theorem 3 determines it after rational coefficients are allowed. In Stanley’s Theorem 3.1 the alphabet length varies through , with independent and in the coefficient field [15]. Here is a fixed integer and the zeros in the variable determine the invariant factors. Chen, Guo, Jin and Liu studied the integer Carlitz matrices and used skew Schur polynomials to relate minors of Toeplitz matrices [2]. Huang and Yang’s path formula and interval decompositions [6 §§4.2, 4.4] give a representation-theoretic method for computing ranks at a fixed root of unity. The length- path between grades and on the tensor product with interval lengths and gives our transposed matrix up to row and column units. For , the exponent adjustment from their to is .
Kuperberg’s question. In each cyclic case of Theorem 1, the group is . Choosing a side of length two for gives a two-row presentation. Lemma 7 gives the first integer Smith factor , and the determinant gives the second, . For any box, Theorem 6 determines the entire -primary group whenever at most two of the polynomials , , divide . With three or more such indices, we still know the number of cyclic -power summands. Corollary 5 supplies a common divisor of the minors of at each intermediate size . The exact power of in the gcd at each size would determine the orders of the individual summands.
Verification
Another AI model independently audited the mathematics and exposition of the 19 November 2026 draft. It found no gap in the proofs and used its own code to compute the integer Smith forms for all 364 boxes with $1 \le a \le b \le c \le 12$.
References
- [1]H. Brenner and A. Kaid, A note on the weak Lefschetz property of monomial complete intersections in positive characteristic. arXiv:1003.0824.arxiv.org/abs/1003.0824
- [2]C. Chen, A. Guo, X. Jin and G. Liu, Trivariate monomial complete intersections and plane partitions, arXiv:1108.1426.
- [3]N. Conlon, On plane partitions and Kasteleyn cokernels, UC Davis REU report, January 2022. Archived report.
- [4]C. Han, The Hilbert–Kunz function of a diagonal hypersurface, Ph.D. thesis, Brandeis University, 1991.
- [5]A. Hazi, Existence and rotatability of the two-colored Jones–Wenzl projector. doi:10.1112/blms.12983.
- [6]H.-L. Huang and Y. Yang, The Green rings of minimal Hopf quivers. doi:10.1017/S0013091515000085.
- [7]H. Joris, C. Oestreicher and J. Steinig, The greatest common divisor of certain sets of binomial coefficients, J. Number Theory 21 (1985), 101–119. doi:10.1016/0022-314X(85)90013-7.
- [8]G. Kuperberg, An exploration of the permanent-determinant method, Electron. J. Combin. 5 (1998), R46. doi:10.37236/1384.DOI
- [9]G. Kuperberg, Kasteleyn cokernels, Electron. J. Combin. 9 (2002), R29. doi:10.37236/1645.DOI
- [10]J. Li and F. Zanello, Monomial complete intersections, the weak Lefschetz property and plane partitions. arXiv:1002.4400.arxiv.org/abs/1002.4400
- [11]P. A. MacMahon, Combinatory Analysis, vol. II, §429, 1916, pp. 182–183. Original text.
- [12]D. Mirmohades, About categorification of cyclotomic integers and tensored N-complexes. arXiv:1807.02100.arxiv.org/abs/1807.02100
- [13]P. Monsky, Mason’s theorem and syzygy gaps, J. Algebra 303 (2006), 373–381.DOI
- [14]J. Propp, Enumeration of matchings: problems and progress, New Perspectives in Algebraic Combinatorics (1999), 255–291. arXiv:math/9904150.arxiv.org/abs/math/9904150
- [15]R. P. Stanley, The Smith normal form of a specialized Jacobi–Trudi matrix, European J. Combin. 62 (2017), 178–182. doi:10.1016/j.ejc.2016.12.008.DOI


