The characteristic polynomial of sunflowers

Changjiang Bu; Ge Lin; Lixiang Chen

REVIEW 3 major objections 4 minor 26 references

The characteristic polynomial of sunflowers

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Sunflower hypergraphs gain explicit characteristic polynomials

desk verdict Genuine extension to non-linear sunflowers with a checkable eigenvector proof, but the main formula as written is not a polynomial and the spectral-radius corollary is wrong; both are fixable. read the letter →

arxiv 2506.17628 v1 pith:PFPHAPKG submitted 2025-06-21 math.CO

Changjiang Bu , Lixiang Chen , Ge Lin This is my paper · ORCID

classification math.CO MSC 05C5005C65

keywords characteristicpolynomialsunflowerhypergraphuniformeigenvaluesspectralmomentsresultantEulerianmulti-digraphalgebraicmultiplicity

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Sunflowers—uniform hypergraphs whose every edge shares the same $s$ seed vertices—form one of the simplest families that are not linear, yet until now their spectra were known only for the single-seed case of hyperstars. This paper establishes a complete answer: for a $k$-uniform sunflower with $s$ seeds and $p$ petals it determines every eigenvalue, every algebraic multiplicity, and every spectral moment, packaged into an explicit product formula for the characteristic polynomial. The result matters because computing the characteristic polynomial of a hypergraph is NP-hard in general, and explicit formulas exist only for a handful of families; sunflowers provide a natural test bed where the non-linear overlap of edges is controlled.

What carries the argument

The argument runs on two coupled engines. First, an eigenvector reduction: writing $\beta$ for the common value determined by $\lambda^{k-s}\beta^s=(x_S)^k$, each petal contributes $x_S x_{P_i}=\beta\xi_i$ with $\xi_i^{s+1}=\xi_i$, and the seed equations force $\lambda^k=(e_p^\top \xi)^s$; this turns eigenvalues into the factors of the product. Second, to count multiplicities, the paper invokes the spectral-moment formula (Lemma 2.1), which expresses $S_d(S)$ as a weighted count of Eulerian multi-digraphs built from rooted hyperedges. The structural lemma (Proposition 3.4) characterizes every such digraph as $D(m,Q)$: a complete multi-digraph of multiplicity $d/k$ on the $s$ seeds, complete multi-digraphs $m_i K_{k-s}$ on each petal, and arcs from seeds to petals with multiplicities $q_{vi}$, subject to the balance equations $e_p^\top m=d/k$, $Qe_p=(d/k)e_s$, $Q^\top e_s=sm$. Matrix-tree and Schur-complement evaluations give the spanning-tree count, and inclusion-exclusion converts the totals into the multiplicity formula.

What would settle it

Compute the characteristic polynomial of a small sunflower such as $S(3,2,2)$ directly from the defining resultant by computer algebra and compare it term by term with the product in Theorem 3.7; a single mismatched coefficient, or an eigenvalue not of the form $\lambda^k=(e_p^\top\xi)^s$, would refute the formula.

Watch

Extended reading notes

Core claim

For $k\ge 3$, the paper shows that a complex number $\lambda$ is an eigenvalue of $S(k,s,p)$ exactly when $\lambda^k = (e_p^\top \xi)^s$ for some $p$-tuple $\xi$ with each $\xi_i^{s+1}=\xi_i$ (so each coordinate is $0$ or an $s$-th root of unity), with the extra restriction that when $s=k-1$ only tuples whose support is empty or all of $[p]$ occur. The characteristic polynomial is then the product of these factors $\lambda^k-(e_p^\top \xi)^s$ raised to multiplicities $\mu(\xi)$: $\mu(0)$ is a closed expression in $k,s,p$, and for $\xi\ne 0$ it is $\frac{1}{s}K^{p-|\mathrm{supp}\,\xi|}k^{|\mathrm{supp}\,\xi|(k-s-1)+s-1}$, where $K=(k-1)^{k-s}-s k^{k-s-1}$. Equivalently, the paper computes the $d$-th spectral moment as a single sum over these $\xi$ of $(e_p^\top \xi)^{sd/k}$ times combinatorial weights, and the moment vanishes unless $k\mid d$. For $k=2$ the formula reduces to the classical star-graph spectrum $\{\pm\sqrt{p},\,0^{p-1}\}$.

Load-bearing premise

The entire multiplicity computation rests on the quoted, unproved spectral-moment formula in Lemma 2.1; if that formula is wrong, every spectral moment and the final characteristic polynomial formula collapses.

Editorial extensions

If this is right

The spectral radius of $S(k,s,p)$ is $\sqrt[k]{ps}$, and its algebraic multiplicity is $k^{p(k-s)+s-1}-p$, confirming the stated conjecture on spectral-radius multiplicity for this family.
The $d$-th order spectral moment vanishes unless $k\mid d$, and otherwise is given by a single explicit sum over the tuples $\xi$; this gives a closed form for all traces of the adjacency tensor of a sunflower.
Setting $s=1$ recovers the known characteristic polynomial of hyperstars $S(k,1,p)$, and setting $k=2$ recovers the star spectrum $\{\pm\sqrt{p},0^{p-1}\}$.
The formula determines all eigenvalues with multiplicities at once, so the spectrum of any sunflower can be read off by evaluating sums over $p$ coordinates chosen from $0$ and the $s$-th roots of unity.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The same $\xi$-parametrization suggests that sunflowers with variable petal sizes or edge weights would still have eigenvalues governed by $\lambda^k=\sum_i w_i\xi_i$ for suitable weights $w_i$; the spectral reduction should survive, while the counting of $D(m,Q)$ would need a generalized version.
Because the multiplicity formula separates by support size, for large $p$ the full-support tuples dominate the algebraic multiplicities; one could use this to approximate the distribution of eigenvalues of large sunflowers.
A natural testable extension is to check whether other symmetric non-linear hypergraphs (for example sunflowers with a shared core that is not a single set) admit the same product structure, which would indicate that the cored-hypergraph symmetry, rather than the sunflower shape specifically, drives the factorization.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

The short version: this paper genuinely extends the known hyperstar case S(k,1,p) to sunflowers S(k,s,p) with s≥2, and the eigenvector construction and spectral moment derivation are mostly sound. But the main theorem is not stated as a polynomial—it uses rational exponents and never groups equal factors—and the spectral-radius corollary is false as written. I think the central result can be fixed, but it needs real revision.

What is new: prior work covered S(k,1,p); this gives eigenvalues, spectral moments, and a candidate characteristic polynomial for the non-linear sunflower family. Theorem 3.1's eigenvalue characterization via Ξ_p = {ξ : ξ_i^{s+1}=ξ_i} is clean, and the constructive eigenvector proof checks out. The digraph characterization in Proposition 3.4 and the matrix-tree computation are careful and explicit; the spectral moment formula in Theorem 3.6 passes small-case checks. That credit is earned.

Where it falls down: Theorem 3.7 writes φ_S(λ) = ∏ (λ^k - (e^T ξ)^s)^{μ(ξ)} with μ(ξ) rational. As a product over individual ξ this is not a polynomial—for k=3, s=2, p=4 each nonzero ξ has μ=3/2. The proof only checks total degree and power sums; that does not establish that the grouped multiplicities are integers. The fix is straightforward: group the product by the value of (e^T ξ)^s and prove the accumulated exponents are nonnegative integers, using the fact that the s-th roots act transitively on coordinates with μ constant on orbits and (e^T ξ)^s invariant. That argument is absent. Separately, Corollary 3.8 states the spectral radius is (ps)^{1/k}; the correct value is p^{s/k}, and the claimed algebraic multiplicity does not match the theorem. The corollary should be corrected or dropped.

This paper is for spectral hypergraph theorists. It deserves serious refereeing—the core method is sound and the result is new—but the revision must restate Theorem 3.7 in grouped polynomial form and fix the corollary. I would send it to review, not accept it as is.

Referee Report

3 major / 4 minor

Summary. The paper studies k-uniform sunflowers S(k,s,p), a family of non-linear uniform hypergraphs, and claims to determine all eigenvalues (Theorem 3.1), all spectral moments (Theorem 3.6), and an explicit characteristic polynomial (Theorem 3.7), together with a corollary on the spectral radius and its algebraic multiplicity (Corollary 3.8). The methods are tensor eigenvalue equations, a Harary–Sachs type spectral-moment formula from the literature, a structural characterization of the relevant Eulerian multi-digraphs, and matrix-tree computations. The eigenvalue construction and the digraph characterization are carefully argued and checkable, and small cases are consistent with Theorems 3.1 and 3.6. However, the central formula in Theorem 3.7 is not a well-defined polynomial as stated, and the proof of the spectral-moment formula contains a counting step that needs correction.

Significance. If the main theorem is repaired and restated correctly, this would be a valuable contribution: it would provide the first explicit characteristic polynomial for a family of non-linear uniform hypergraphs, confirm a conjecture of Fan for sunflowers, and demonstrate a transferable combination of spectral moments and matrix-tree methods. The eigenvalue construction in Theorem 3.1 and the Eulerian digraph characterization in Proposition 3.4 are specific and checkable. The paper does not include machine-checked proofs or code, but the derivations are explicit enough to be verified on small examples.

major comments (3)

[Theorem 3.7] The central formula is not a well-defined characteristic polynomial as written. The exponents μ(ξ) are rational in general: for k=3, s=2, p=4, every full-support ξ in {±1}^4 has μ(ξ)=3/2, and for k=3, s=1, p=2, μ(0)=35/3. A product of powers of polynomials with rational exponents is not an element of C[λ], and the proof does not establish that, after collecting factors with equal (e_p^T ξ)^s, the exponents become nonnegative integers. The verification of total degree and of the power sums S_d(S) in the proof of Theorem 3.7 only shows that the rational-exponent multiset has the correct moments; it does not by itself certify polynomiality. The theorem should be restated in grouped form with integer exponents, and the integrality argument (for example via the μ_s-action on Ξ_p and the congruence that makes kμ(0) integral) should be supplied.
[Proof of Theorem 3.6, counting of f] The displayed equality |{f : D_f = D(m,Q), constraints}| = ((k-1)!)^d (s d/k)! / ∏_i (s m_i)! is false for a fixed multi-digraph D(m,Q). For fixed Q the correct count is ((k-1)!)^d ∏_{v∈S} (d/k)! / ∏_i q_{v,i}!, which depends on Q. For example, with k=3, s=2, p=2, d=6 and m=(1,1), the matrices Q=[[2,0],[0,2]] and Q=[[1,1],[1,1]] give counts 64 and 256, respectively, not 384 each. The displayed expression ϕ(m) is correct only after summing over all Q satisfying the constraints, via the identity Σ_Q ∏_{v∈S} (d/k)!/∏_i q_{v,i}! = (s d/k)! / ∏_i (s m_i)!. The proof should make this Q-summation explicit; as written, it contains an incorrect intermediate statement in a load-bearing step of the spectral-moment computation.
[Corollary 3.8] The spectral radius stated as k√ps is inconsistent with Theorems 3.1 and 3.7, which imply the largest eigenvalue is (p^s)^{1/k}. For k=3, s=2, p=4, Theorem 3.1 gives λ^3 = 16, so the spectral radius is 2^{4/3}, not 2. The algebraic multiplicity formula should also be written unambiguously as k^{p(k-s)+s-1-p} (equivalently k^{p(k-s-1)+s-1}); in the present plain-text rendering it is ambiguous. Please correct the value and clarify the notation.

minor comments (4)

[Theorem 3.7 vs Theorem 3.1] The product in Theorem 3.7 ranges over all ξ∈Ξ_p, but for s=k-1 the proof sets μ(ξ)=0 for ξ∉Ξ_p^0; stating this explicitly in the theorem statement would avoid confusion.
[Proof of Theorem 3.7, first paragraph] The sentence claiming that the form of φ(λ) 'follows' from Theorem 3.1 and k-symmetry is too quick; the authors should explain how the spectral moments determine the multiplicities uniquely after the factors are grouped by equal (e_p^T ξ)^s.
[Corollary 3.8, notation] The notation k√ps is ambiguous; please use √[k]{p^s} and write the exponent explicitly as k^{ p(k-s)+s-1-p}.
[Definition 3.3 and preceding paragraph] The arrow notation V --m--> U is defined just before Lemma 3.2 but used again in Definition 3.3; re-stating it at the point of use would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the sunflower characteristic polynomial derivation is self-contained apart from a standard quoted spectral-moment formula.

full rationale

The paper's central derivation is not circular. Theorem 3.1 independently characterizes the eigenvalues of a sunflower by explicitly constructing eigenvectors from solutions of the defining tensor equations, and Theorem 3.6 computes spectral moments from the general moment formula Lemma 2.1 quoted from Chen, van Dam, and Bu [6]. Although two authors of the present paper overlap with [6], Lemma 2.1 is a published, externally reviewed general result and is not the target of this paper; it is not equivalent to the sunflower characteristic polynomial by construction. Theorem 3.7 then proposes an explicit candidate polynomial and verifies it by checking both the total degree and the coincidence of all d-th power sums with the independently computed spectral moments S_d(S), which is a standard sufficient test for equality of polynomials when the candidate has the correct root support. No parameter is fitted to the output and no prediction is renamed from an input. The rational exponents in Theorem 3.7 and the unproved integrality of the grouped multiplicities are well-formedness and correctness concerns, not instances of circular reasoning, since the proof does not assume the conclusion. The derivation therefore receives score 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new free parameters or postulated entities. It relies on two domain lemmas from the prior literature (the spectral moment formula and k-symmetry of cored hypergraphs) and on standard algebraic tools (resultant degree, Matrix-Tree theorem, Schur complements).

assumptions (4)

domain assumption Lemma 2.1, the formula for the d-th order spectral moment of a k-uniform hypergraph in terms of Eulerian directed graphs, is correct and applicable.
Quoted from [6] without proof; it is the foundation for all spectral moment computations in Section 3.2.
domain assumption The spectrum of a cored k-uniform hypergraph is k-symmetric, so S_d(H)=0 whenever k does not divide d.
Used at the start of Section 3.2 to restrict attention to k|d; cited from Shao, Qi, and Hu [24].
standard math The characteristic polynomial of a k-uniform hypergraph on n vertices has degree n(k-1)^(n-1).
Standard resultant degree formula, used in the proof of Theorem 3.7 to determine μ(0).
standard math The Matrix-Tree theorem and the Schur complement determinant formula correctly compute the number of spanning trees of the directed graphs D(m,Q).
Used in the proof of Lemma 3.5(b) to derive t(D).

how reviews work

0 comments

Cite this review

Pith. "Pith review of The characteristic polynomial of sunflowers." pith.science (2026). https://pith.science/paper/PFPHAPKG

@misc{pith2026250617628,
  author       = {Pith},
  title        = {Pith review of: The characteristic polynomial of sunflowers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PFPHAPKG}},
  note         = {Machine review of arXiv:2506.17628}
}

read the original abstract

A uniform hypergraph is called a sunflower if all of its hyperedges intersect in the same set of vertices. In this paper, we determine the eigenvalues and spectral moments of a sunflower, thereby obtaining an explicit formula for its characteristic polynomial.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 25 canonical work pages

[1]

Y. Bao, Y. Fan, Y. Wang, and M. Zhu. A combinatorial method for comput- ing characteristic polynomials of starlike hypergraphs.Journal of Algebraic Combinatorics, 51(4):589–616, 2020. 14

work page 2020
[2]

Brualdi and H

R. Brualdi and H. Schneider. Determinantal identities: Gauss, Schur, Cauchy, Sylvester, Kronecker, Jacobi, Binet, Laplace, Muir, and Cayley. Linear Algebra and its Applications, 52:769–791, 1983

work page 1983
[3]

Chang, K

K. Chang, K. Pearson, and T. Zhang. Perron-Frobenius theorem for non- negative tensors.Communications in Mathematical Sciences, 6(2):507–520, 2008

work page 2008
[4]

Chen and C

L. Chen and C. Bu. A reduction formula for the characteristic polynomial of hypergraph with pendant edges.Linear Algebra and its Applications, 611:171–186, 2021

work page 2021
[5]

Chen and C

L. Chen and C. Bu. The algebraic multiplicity of the spectral radius of a uniform hypertree.The Electronic Journal of Combinatorics, 31(4), 2024. Article ID # P4.18

work page 2024
[6]

Chen, E.R

L. Chen, E.R. van Dam, and C. Bu. Spectra of power hypergraphs and signed graphs via parity-closed walks.Journal of Combinatorial Theory, Series A, 207:105909, 2024

work page 2024
[7]

Clark and J

G. Clark and J. Cooper. A Harary-Sachs theorem for hypergraphs.Journal of Combinatorial Theory, Series B, 149:1–15, 2021

work page 2021
[8]

Cooper and A

J. Cooper and A. Dutle. Spectra of uniform hypergraphs.Linear Algebra and its Applications, 436(9):3268–3292, 2012

work page 2012

Show all 26 references

[9]

Cooper and A

J. Cooper and A. Dutle. Computing hypermatrix spectra with the Poisson product formula.Linear and Multilinear Algebra, 63(5):956–970, 2015

2015
[10]

D. Cox, J. Little, and D. O’Shea.Using algebraic geometry. Springer, New York, 2005

2005
[11]

Cvetkovi´ c, M

D. Cvetkovi´ c, M. Doob, and H. Sachs.Spectra of Graphs: Theory and Applications. Academic Press, New York, 1980

1980
[12]

Duval, C

A. Duval, C. Klivans, and J. Martin. Simplicial matrix-tree theorems. Transactions of the American Mathematical Society, 361(11):6073–6114, 2009

2009
[13]

Y. Fan. The multiplicity of eigenvalues of nonnegative weakly irreducible tensors and uniform hypergraphs. arXiv:2410.20830v2, 2024

2024 arXiv
[14]

Hillar and L

C. Hillar and L. Lim. Most tensor problems are NP-hard.Journal of the ACM, 60(6):1–39, 2013

2013
[15]

S. Hu, Z. Huang, C. Ling, and L. Qi. On determinants and eigenvalue theory of tensors.Journal of Symbolic Computation, 50:508–531, 2013

2013
[16]

S. Hu, L. Qi, and J. Shao. Cored hypergraphs, power hypergraphs and their Laplacian H-eigenvalues.Linear Algebra and its Applications, 439(10):2980–2998, 2013. 15

2013
[17]

Jouanolou

J.-P. Jouanolou. Le formalisme du r´ esultant.Advances in Mathematics, 90(2):117–263, 1991

1991
[18]

H. Li, L. Su, and S. Fallat. On a relationship between the characteristic and matching polynomials of a uniform hypertree.Discrete Mathematics, 347(5):113915, 2024

2024
[19]

L. Lim. Singular values and eigenvalues of tensors: a variational approach. In1st IEEE International Workshop on Computational Advances in Multi- Sensor Adaptive Processing, pages 129–132. IEEE, 2005

2005
[20]

Macaulay

F. Macaulay. On some formulas in elimination.Proceedings of the London Mathematical Society, 3:3–27, 1902

1902
[21]

Morozov and Sh

A. Morozov and Sh. Shakirov. Analogue of the identity Log Det = Trace Log for resultants.Journal of Geometry and Physics, 61(3):708–726, 2011

2011
[22]

L. Qi. Eigenvalues of a real supersymmetric tensor.Journal of Symbolic Computation, 40(6):1302–1324, 2005

2005
[23]

L. Qi. H +-eigenvalues of Laplacian and signless Laplacian tensors.Com- munications In Mathematical Sciences, 12(6):1045–1064, 2014

2014
[24]

J. Shao, L. Qi, and S. Hu. Some new trace formulas of tensors with ap- plications in spectral hypergraph theory.Linear and Multilinear Algebra, 63(5):971–992, 2015

2015
[25]

Stanley.Enumerative Combinatorics

R. Stanley.Enumerative Combinatorics. Volume 1. Cambridge University Press, Cambridge, 2012

2012
[26]

Y. Zheng. The characteristic polynomial of the complete 3-uniform hyper- graph.Linear Algebra and its Applications, 627:275–286, 2021. 16

2021

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Y. Bao, Y. Fan, Y. Wang, and M. Zhu. A combinatorial method for comput- ing characteristic polynomials of starlike hypergraphs.Journal of Algebraic Combinatorics, 51(4):589–616, 2020. 14

work page 2020

[2] [2]

Brualdi and H

R. Brualdi and H. Schneider. Determinantal identities: Gauss, Schur, Cauchy, Sylvester, Kronecker, Jacobi, Binet, Laplace, Muir, and Cayley. Linear Algebra and its Applications, 52:769–791, 1983

work page 1983

[3] [3]

Chang, K

K. Chang, K. Pearson, and T. Zhang. Perron-Frobenius theorem for non- negative tensors.Communications in Mathematical Sciences, 6(2):507–520, 2008

work page 2008

[4] [4]

Chen and C

L. Chen and C. Bu. A reduction formula for the characteristic polynomial of hypergraph with pendant edges.Linear Algebra and its Applications, 611:171–186, 2021

work page 2021

[5] [5]

Chen and C

L. Chen and C. Bu. The algebraic multiplicity of the spectral radius of a uniform hypertree.The Electronic Journal of Combinatorics, 31(4), 2024. Article ID # P4.18

work page 2024

[6] [6]

Chen, E.R

L. Chen, E.R. van Dam, and C. Bu. Spectra of power hypergraphs and signed graphs via parity-closed walks.Journal of Combinatorial Theory, Series A, 207:105909, 2024

work page 2024

[7] [7]

Clark and J

G. Clark and J. Cooper. A Harary-Sachs theorem for hypergraphs.Journal of Combinatorial Theory, Series B, 149:1–15, 2021

work page 2021

[8] [8]

Cooper and A

J. Cooper and A. Dutle. Spectra of uniform hypergraphs.Linear Algebra and its Applications, 436(9):3268–3292, 2012

work page 2012