REVIEW 1 major objections 7 minor 20 references
On commuting integer matrices
T0 review · 1 major / 7 minor · reviewed 2026-08-16 · deepseek-v4-flash
Pith's one-line read The paper proves that the count of commuting pairs of 3x3 integer matrices in [-N,N]^{3x3} has order N^10, and gives an asymptotic with explicit constant for 2x2 matrices.
desk verdict Sharp count for commuting 3x3 integer matrices and first asymptotic for 2x2; elementary, correct, and worth a serious referee. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The carrying object is the restricted divisor correlation $r_N(h)$, the number of $a_1,a_2,a_3,a_4\in[-N,N]$ with $a_1a_2-a_3a_4=h$, together with its moments $I_k(N)=\sum_h r_N(h)^k$. Lemma 1.5 gives the pointwise bound $r_N(h)\ll N^2\sum_{d\mid h,\,d\le N}1/d$ for $0<|h|\le 2N^2$ and the moment bound $I_k(N)\ll_k N^{2k+2}$. These bounds fix the number of admissible off-diagonal entry choices in the rank-$4$ case, while the rank stratification of the matrix $M$ in (4.3) controls how many diagonal choices remain.
What would settle it
Compute $I_3(N)=\sum_{|h|\le 2N^2} r_N(h)^3$ for increasing $N$ and check whether it stays within a constant of $N^8$; a single $N$ with $I_3(N)>C N^8$ for large $C$ would refute Lemma 1.5. A more refined check is to evaluate the pointwise bound (1.5) for $h$ of size roughly $2N^2$ by direct enumeration of the $O(N^4)$ quadruples at moderately large $N$.
Extended reading notes
Core claim
The central claim is that the naive lower bound is sharp for $3\times 3$ commuting integer matrices: $N^{10}\ll C_3(N)\ll N^{10}$ for every positive integer $N$. For $2\times 2$ matrices the paper goes further and gives the first-order asymptotic $C_2(N)=K(2N)^5+O(N^4\log N)$ with $K=10\zeta(2)/(3\zeta(3))$. The mechanism is to count solutions to $AB=BA$ by first using three quadratic equations to fix the off-diagonal entries and then classifying by the rank of a $6\times 4$ matrix $M$ whose entries are those off-diagonal entries; the rank-$4$ case, which dominates, is controlled by the third moment estimate $I_3(N)=O(N^8)$. A separate local computation over $\mathbb{Z}/p^n\mathbb{Z}$ shows that the local densities multiply to a divergent product, which explains why the circle-method heuristic would predict spurious logarithmic factors.
Load-bearing premise
The load-bearing premise is the moment estimate $I_3(N)=O(N^8)$ for the correlation function $r_N(h)$; if that count of coincident $2\times 2$ determinants were larger, the rank-4 case in the proof of Theorem 1.1 would not fit in $N^{10}$.
Editorial extensions
If this is right
- For $d=3$ the commutator variety's point count has the same order as its trivial subfamily, so the generic matrix pair contributes at most a bounded factor.
- The $2\times 2$ asymptotic removes the implicit $N^\epsilon$ factor and supplies a numerically explicit leading constant $K\approx 4.56144$.
- The $p$-adic density computation shows the local factors do not multiply to a finite constant, so the usual singular series heuristic fails for this problem; the main term instead comes from pairs where one matrix is a scalar multiple of the identity.
- For arbitrary finite sets $A\subset\mathbb{R}$ with small doubling, $C_3(A)$ has order $|A|^{10}$ up to a factor depending only on the doubling constant $K$.
Reading between the lines
- For $d\ge 4$ the same rank stratification cannot suffice on its own: the consistency example in the paper's final remarks shows a full-rank $M$ with empty solution set, so counting $C_d(N)$ will require an additional constraint that has no analogue for $d=3$.
- A weighted analogue of $r_N(h)$, using smooth bump functions instead of sharp truncation, should satisfy an identical moment bound and would likely give the same $N^{10}$ order for smooth box-constrained matrices.
- The explicit constant in the $2\times 2$ asymptotic suggests that higher-dimensional analogues, when they become available, will express the leading coefficient in terms of zeta values and totient sums.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies the number C_d(N) of pairs of d×d integer matrices with entries in [-N,N] that commute. Theorem 1.1 proves the sharp order N^{10} for d=3, confirming a conjecture of Browning–Sawin–Wang. Theorem 1.2 gives an asymptotic formula C_2(N) = K(2N)^5 + O(N^4 log N) with an explicit constant K = 10ζ(2)/(3ζ(3)). Theorem 1.3 provides the analogous local count over Z/p^nZ, and Theorem 1.6 extends the d=3 upper bound to arbitrary finite sets A ⊂ R with small doubling. The proofs are elementary, based on new restricted divisor-correlation estimates (Lemma 1.5).
Significance. The main results are significant. The d=3 order-of-magnitude problem was open, and the d=2 asymptotic with explicit constant is new. The proof method is self-contained and avoids harmonic analysis; the pointwise and moment bounds for r(h) are of independent interest. The paper also gives a clean p-adic treatment and a nice application of Solymosi's sum-product bound. However, the lower-bound half of Theorem 1.6 is not proved in the manuscript.
major comments (1)
- [Section 6.2 (proof of Theorem 1.6)] The proof of Theorem 1.6 as written establishes only the upper bound C_3(A) ≪ K^6 |A|^{10} (log(2|A|))^3. After Lemma 6.1, the argument splits into rank cases and bounds each case from above; there is no lower-bound construction for arbitrary A. The lower bound K^{-3}|A|^{10} ≪ C_3(A) is asserted in the theorem statement but is not obtained from the scalar-multiple construction used in the integer case, which requires 0 ∈ A. The authors should either supply a proof of this lower bound or modify the statement of Theorem 1.6 to include only the upper bound (and any lower bound that does follow from the given arguments).
minor comments (7)
- [Lemma 2.1] The statement reads 'let u, w be coprime, non-zero integers' but should read 'let u, v be coprime, non-zero integers'.
- [Lemma 2.2 proof] In the proof, 'b − b′ = |vz|' should read 'b − b′ = vz'.
- [Proof of Lemma 1.5] The displayed line for r(0) ends with '16 ζ(2)N^2 log N'; the preceding calculation gives 16/ζ(2) as the coefficient. Since the lemma only needs the order N^2 log N, this typo does not affect the results.
- [Proof of Lemma 1.5] The term 'N^{1+1/2}' appears to be a typographical artifact for N^{1+ε}; the subsequent N^2 bound is sufficient regardless.
- [Section 4, after (4.1)] The text says the matrices lie in 'Mat2(Z,N)', but they are 3×3 matrices, so this should be 'Mat3(Z,N)'.
- [Lemma 4.2] In the final display, the bound is written as '|S3| ≪ N^8 · N^2' but the lemma estimates |S4|, so the subscript should be 4.
- [Section 6.3] The phrase 'the the first six rows' contains a duplicated article.
Circularity Check
No significant circularity: the central estimates are proven in-paper and independent of the claims they support.
full rationale
The derivation chain is self-contained. Lemma 1.5, the load-bearing estimate, is proved elementarily in Section 3 from Lemma 3.1, an independent convolution bound that is itself proved from first principles; the proof of the pointwise bound r(h) << N^2 sum_{d|h,d<=N} 1/d fixes a1,a3 and counts a2,a4 by a congruence, and the moment bound I_k(N) << N^{2k+2} is deduced from this pointwise bound and Lemma 3.1, not assumed. Section 4 then uses only these estimates: the rank-4 case is bounded by I_3(N)=O(N^8) times O(N^2) diagonal choices, the rank-3 case by r(0)^3=O(N^6 log^3 N) times O(N^3), and the rank-2 cases by direct counting of parallel direction vectors. Theorem 1.2 is obtained by exact counting in Section 2 using Lemmas 2.1 and 2.2 together with standard Euler-totient estimates; no fitted parameter is later renamed as a prediction. Theorem 1.3 is likewise proved by an independent p-adic lifting argument (Lemma 5.1). The only self-citation, to the second author's [17], appears in introductory context or in the arbitrary-set Theorem 1.6, whose proof invokes Solymosi's external sum-product bound [20] and the paper's own Section 4 scheme; it is not load-bearing for Theorems 1.1 or 1.2. No equation in the paper reduces to its own conclusion by construction, and the paper explicitly marks the d>=4 discussion as illustrative rather than a claimed theorem.
Assumptions & free parameters
assumptions (3)
- standard math Euler totient summatory estimates: sum_{u<=N} phi(u)/u^3 = zeta(2)/zeta(3) - 1 + O(1/N) and sum_{u<=N} phi(u)/u^2 = log N / zeta(2) + O(1).
- standard math Divisor function bound tau(n) <<_epsilon n^epsilon.
- standard math Solymosi's sum-product bound: for a finite set A with |A+A| = K|A|, the multiplicative energy r_A(0) is O(K^2 |A|^2 log(2|A|)).
Cite this review
Pith. "Pith review of On commuting integer matrices." pith.science (2026). https://pith.science/paper/JBXCHY7V
@misc{pith2026250415839,
author = {Pith},
title = {Pith review of: On commuting integer matrices},
year = {2026},
howpublished = {\url{https://pith.science/paper/JBXCHY7V}},
note = {Machine review of arXiv:2504.15839}
}
abstract
Given $d, N \in \mathbb{N}$, we define $\mathfrak{C}_d(N)$ to be the number of pairs of $d\times d$ matrices $A,B$ with entries in $[-N,N] \cap \mathbb{Z}$ such that $AB = BA$. We prove that $$ N^{10} \ll \mathfrak{C}_3(N) \ll N^{10},$$ thus confirming a speculation of Browning-Sawin-Wang. We further establish that $$ \mathfrak{C}_2(N) = K(2N+1)^5 (1 + o(1)),$$ where $K>0$ is an explicit constant. Our methods are completely elementary and rely on upper bounds of the correct order for restricted divisor correlations with high uniformity.
Reference graph
Works this paper leans on
-
[4]
T. Browning, W. Sawin, V. Y. Wang, Pairs of commuting integer matrices , arXiv:2409.01920
-
[17]
Mudgal, On commuting pairs in arbitrary sets of 2 × 2 matrices, arXiv:2411.10404
A. Mudgal, On commuting pairs in arbitrary sets of 2 × 2 matrices, arXiv:2411.10404
-
[1]
M. Afifurrahman, A uniform formula on the number of integer matrices with give n determinant and height, arXiv:2407.08191
-
[2]
T. Apostol, Introduction to analytic number theory , Undergraduate Texts in Mathematics, Springer- Verlag, New York-Heidelberg, 1976
work page 1976
-
[3]
Basili, On the irreducibility of varieties of commuting matrices , J
R. Basili, On the irreducibility of varieties of commuting matrices , J. Pure Appl. Algebra 149 (2000), no. 2, 107–120
work page 2000
-
[5]
W. Duke, Z. Rudnick, P. Sarnak, Density of integer points on affine homogeneous varieties , Duke Math. J. 71 (1993) 143–179
work page 1993
-
[6]
Erd˝ os, P
P. Erd˝ os, P. Tur´ an,On some problems of a statistical group-theory. IV , Acta Math. Acad. Sci. Hungar. 19 (1968), 413–435
1968
-
[7]
W. Feit, N. J. Fine, Pairs of commuting matrices over a finite field , Duke Math. J. 27 (1960), 91-94
1960
Show all 20 references
-
[8]
Ganguly, R
S. Ganguly, R. Guria, Lattice points on determinant surfaces and the spectrum of t he automorphic Laplacian, arXiv:2410.04637
-
[9]
W. H. Gustafson, What is the probability that two group elements commute? , Amer. Math. Monthly 80 (1973), 1031–1034
1973
-
[10]
D. R. Heath-Brown, The fourth power moment of the Riemann zeta function , Proc. London Math. Soc. (3) 38 (1979), no. 3, 385–422
1979
-
[11]
D. R. Heath-Brown, A new form of the circle method, and its application to quadra tic forms , J. Reine Angew. Math. 481 (1996), 149–206
1996
-
[12]
C. R. Johnson, M. G. Marques, Patterns of commutativity: the commutant of the full patter n, Electron. J. Linear Algebra 14 (2005), 43–50
2005
-
[13]
C. R. Johnson, O. Walch, Commuting pairs of patterns and symmetric realizations , Electron. J. Linear Algebra 25 (2012), 84–91
2012
-
[14]
Tom Meurman, On the binary additive divisor problem , Number theory (Turku, 1999), 223—246, Walter de Gruyter & Co., Berlin, 2001
1999
-
[15]
Motohashi, The binary additive divisor problem , Ann
Y. Motohashi, The binary additive divisor problem , Ann. Sci. ´Ecole Norm. Sup. (4) 27 (1994), no. 5, 529–572
1994
-
[16]
T. S. Motzkin, O. Taussky, Pairs of matrices with property L. II. Trans. Amer. Math. Soc. 80 (1955), 387–401
1955
-
[18]
P. M. Neumann, Two combinatorial problems in group theory , Bull. London Math. Soc. 21 (1989), no. 5, 456–458
1989
-
[19]
Salberger, Counting rational points on projective varieties , Proc
P. Salberger, Counting rational points on projective varieties , Proc. Lond. Math. Soc. (3) 126 (2023), no. 4, 1092–1133
2023
-
[20]
Solymosi, Bounding multiplicative energy by the sumset , Adv
J. Solymosi, Bounding multiplicative energy by the sumset , Adv. Math. 222 (2009), no. 2, 402-408. Mathematics Institute, Zeeman Building, University of W ar wick, Coventry CV4 7AL, United Kingdom Email address : Jonathan.Chapman@warwick.ac.uk Mathematics Institute, Zeeman Bui...
2009
Reviewed August 16, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.