REVIEW 2 major objections 4 minor 16 references
On the signless Laplacian spectral radius of $K_{s,t}$-minor free graphs
T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read For $K_{2,t}$-minor-free graphs of large order, the signless Laplacian spectral radius obeys a sharp closed-form upper bound, with equality exactly on the join graph $F_{2,t}(n)$; the $K_{3,3}$-minor-free case is also settled for $n\ge…
desk verdict Sharp signless Laplacian bounds for K2,3- and K3,3-minor-free graphs, with the K3,3 case resting on four unproved adapted lemmas that need proof or exact reference. 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 load-bearing family is $F_{s,t}(n)=K_{s-1}\vee(p\cdot K_t\cup K_r)$, where $n-s+1=pt+r$ and $0\le r<t$: a clique of $s-1$ universal vertices joined to disjoint copies of $K_t$ and one $K_r$. The proof first shows that a spectral extremal graph must be edge-maximal, then uses Perron–Frobenius and a degree-weighted inequality to force one universal vertex for $K_{2,t}$ or two universal vertices for $K_{3,3}$. Once those universal vertices exist, the remaining graph must consist of disjoint triangles and at most a path of length 1 or 2, and the paper proves by explicit edge switches that any longer path would raise $q(G)$, contradicting maximality. Four adapted degree-sequence lemmas carry the delicate step of capping $q$ at $n+2$ for graphs with near-universal but not universal vertices.
What would settle it
Enumerate all degree sequences satisfying the hypotheses of Lemmas 2.5–2.8 near the thresholds and compute the signless Laplacian spectral radius; a single sequence with $q(G)>n+2$ would break the key bound that forces the $K_{3,3}$ extremal graph to contain two universal vertices. Alternatively, build an edge-maximal $K_{3,3}$-minor-free graph with two universal vertices and one remaining path of length 3; the proof's switching argument predicts it cannot be extremal, so any such graph with $q$ exceeding $q(F_{3,3}(n))$ would refute Theorem 1.4.
Extended reading notes
Core claim
On its own terms, the discovery is an extremal classification. If $G$ has no $K_{2,t}$ minor and order $n\ge t^2+4t+1$ with $t\ge3$, then $q(G)\le \frac{n+2t-2+\sqrt{(n-2t+2)^2+8t-8}}{2}$, with equality precisely when $n\equiv1\pmod t$ and $G=F_{2,t}(n)$. For $t=3$ this sharpens to the statement that for $n\ge22$, $F_{2,3}(n)$ is the unique maximizer. For $K_{3,3}$-minor-free graphs, the paper shows that for $n\ge1186$ the unique maximizer is $F_{3,3}(n)$, with $q(F_{3,3}(n))$ the largest root of the displayed cubic. The paper also states what it calls Conjecture 4.4: for $2\le s\le t$ and sufficiently large $n$, $F_{s,t}(n)$ should be the unique maximizer among $K_{s,t}$-minor-free graphs.
Load-bearing premise
The $K_{3,3}$ proof depends on four degree-sequence lemmas taken from earlier work in slightly changed forms; the paper says these altered forms are correct but does not prove them, and if any one fails, the conclusion that the extremal graph must have two universal vertices collapses.
Editorial extensions
If this is right
- For $K_{2,3}$-minor-free graphs with $n\ge22$, the maximum signless Laplacian spectral radius is exactly $q(F_{2,3}(n))$, the largest root of the cubic in Lemma 2.3(ii), and $F_{2,3}(n)$ is the unique extremal graph.
- For $t\ge4$, the upper bound is tight exactly when $n\equiv1\pmod t$; in other residue classes the inequality is strict, so the theorem leaves the exact maximum open for those orders.
- For $K_{3,3}$-minor-free graphs of order $n\ge1186$, $F_{3,3}(n)$ is the unique extremal graph, completely settling the $Q$-spectral extremal problem for this minor class at large orders.
- The extremal graphs in all settled cases are edge-maximal and contain one or two universal vertices, so any future counterexample would have to avoid that structure.
- The paper's Conjecture 4.4 says the same join family $F_{s,t}(n)$ should maximize $q$ among all $K_{s,t}$-minor-free graphs for $2\le s\le t$ and sufficiently large $n$.
Reading between the lines
- The residue condition $n\equiv1\pmod t$ looks like a technical artifact: for orders in other residue classes the stated bound is not attained, and the natural prediction is that $F_{2,t}(n)$ still maximizes $q$, merely with a value strictly below the bound.
- The explicit edge-switch arguments suggest a general smoothing principle for the signless Laplacian: in spectral extremal minor-free graphs, long pendant paths are unstable, so extremal graphs should always be highly clustered unions of cliques around a small universal core.
- The four adapted degree-sequence lemmas are the fragile hinge of the $K_{3,3}$ proof; checking them computationally on degree sequences near $n=1186$ would be a cheap way to test whether the structural conclusion is sound.
- If the same strategy is pushed to general $K_{s,t}$, the thresholds should grow polynomially in $t$ rather than exponentially, and the current bounds $t^2+4t+1$ and $1186$ are the natural starting points.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies the signless Laplacian spectral radius q(G) of K_{s,t}-minor-free graphs. Its main results are: (i) for K_{2,3}-minor-free graphs of order n≥22, q(G) ≤ q(F_{2,3}(n)) with equality if and only if G=F_{2,3}(n); (ii) for t≥4 and K_{2,t}-minor-free graphs of order n≥t^2+4t+1, q(G) ≤ (n+2t−2+√((n−2t+2)^2+8t−8))/2 with equality if and only if n≡1 (mod t) and G=F_{2,t}(n); and (iii) for K_{3,3}-minor-free graphs of order n≥1186, q(G) ≤ q(F_{3,3}(n)) with equality if and only if G=F_{3,3}(n). The proofs combine spectral perturbation arguments with known edge-density and degree-sequence results. The paper also proposes Conjecture 4.4 for general K_{s,t}-minor-free graphs of large order.
Significance. If the missing justifications are supplied, the results constitute a meaningful contribution to spectral extremal graph theory for minor-closed graph classes. Theorems 1.2 and 1.3 are essentially self-contained modulo cited lemmas and give sharp, explicit upper bounds together with unique extremal graphs. The paper makes good use of Perron-Frobenius theory, Rayleigh quotient inequalities, and known edge-density bounds. The main caveat is that Theorem 1.4 depends on four degree-sequence lemmas (Lemmas 2.5–2.8) that are stated in adapted form without proof or a precise reference; if those lemmas are correct, the K_{3,3} result is plausible and well supported otherwise.
major comments (2)
- [Section 2 (Lemmas 2.5–2.8)] The paper states that Lemmas 2.5–2.8 are 'a little different from their original forms [16], but indeed they are correct according to original proofs,' yet it neither proves them nor identifies the exact statements in [16] from which they follow. These lemmas are load-bearing for Theorem 1.4: Lemma 4.2 Case 1 invokes Lemmas 2.5 and 2.6, and Lemma 4.3 Case 1 invokes Lemmas 2.7 and 2.8, to obtain the crucial bound q(G) ≤ n+2. If any of the adapted lemmas fails, the proof cannot force Δ(G)=Δ′(G)=n−1, and the uniqueness of F_{3,3}(n) collapses. The adaptation is nontrivial because [16] concerns planar graphs with e(G) ≤ 3n−6, whereas K_{3,3}-minor-free graphs can have e(G) ≤ 3n−5; the changed constants and ranges are essential to the edge-count inequalities in Lemmas 4.2 and 4.3. The authors must provide full proofs of the adapted lemmas or a detailed derivation from the original results.
- [Section 4 (Theorem 1.4 Claim, Cases 1–3; also Theorem 1.2 Claim)] The proofs repeatedly assert 'Clearly G′ is K2,3-minor free' or 'Clearly G′ is K3,3-minor free' after adding edges to path endpoints or performing edge switches (e.g., Theorem 1.2 Cases 1–3, Theorem 1.4 Cases 1–3). These assertions are load-bearing, because the contradiction relies on G′ being minor-free and having a larger signless Laplacian spectral radius. The claims are likely correct, but they require at least a short argument showing that the described operations cannot create the forbidden minor from the path/triangle structure. The authors should supply these justifications rather than leaving them to the reader.
minor comments (4)
- [Proof of Theorem 1.4] The text contains a typo: 'Perron–Fronbenius' should be 'Perron–Frobenius'.
- [Proofs of Theorems 1.2 and 1.3] There are minor language errors: 'Furtherer' in Theorem 1.2 should be 'Further', and 'combing' in Theorem 1.3 should be 'combining'.
- [Lemma 4.1] In the displayed inequality, the parentheses are unbalanced: the term should read '(e(G) − d(u) − e(N(u)))' with a closing parenthesis before the equal sign.
- [Lemmas 2.5 and 2.7] The notation 'n/6 + 1 ≤ d_{k+1} ≤ · · · ≤ d_2 ≤ n − 61' is confusing because degree sequences are nonincreasing; rewriting the hypotheses as d_2 ≤ n−61 and d_{k+1} ≥ n/6+1 would improve readability.
Circularity Check
No significant circularity: all load-bearing results are external theorems or direct spectral computations, not self-citations or fitted inputs.
full rationale
The derivation chain is self-contained in the relevant sense. Theorem 1.2 builds on Lemma 2.1, Theorem 1.3 builds on Lemma 2.2, and Theorem 1.4 builds on Lemmas 2.5-2.8 and 4.1-4.3; all of these are either proved in the paper or cited from prior work by non-overlapping authors, so there is no self-citation chain carrying the argument. The extremal candidates F_{s,t}(n) are defined independently and their signless Laplacian spectral radii are computed directly in Lemma 2.3 rather than being fitted from the target bound. No parameter is calibrated to a subset of the data, no quantity that is meant to be predicted is used as an input, and no uniqueness claim is imported from the authors' own prior work as an external fact. The paper's statement that Lemmas 2.5-2.8 are 'a little different from their original forms' but correct is a possible proof-completeness gap, not circularity: the lemmas come from an external source, and the paper does not redefine its conclusion in terms of them. If the adaptations are incorrect, the proof of Theorem 1.4 would fail, but that is a correctness risk rather than a circularity, and the hard rules require an exhibited reduction of the result to its own inputs, which is not present here.
Assumptions & free parameters
assumptions (7)
- standard math Perron-Frobenius theorem for irreducible nonnegative matrices
- domain assumption Lemma 2.2 from [8]: K2,t-free graphs of order n ≥ t^2+4t+1 satisfy the stated q bound, with equality iff G = K1 ∨ H for (t-1)-regular H
- domain assumption Lemma 2.1 from [8]: K2,3-free graphs of order n ≥ 22 with maximum degree at most n-2 have q < n
- domain assumption Lemma 2.4 from [5]: K1,t-minor free graphs of order n ≥ t+2 have at most n + t(t-3)/2 edges
- domain assumption Lemmas 2.5 through 2.8 from [16] as modified by the authors
- domain assumption Edge bound e(G) ≤ 3n-5 for K3,3-minor free graphs from [6]
- domain assumption Edge bound e(N(u)) ≤ 2d(u)-2 for K2,3-minor free induced subgraphs from [4]
Cite this review
Pith. "Pith review of On the signless Laplacian spectral radius of $K_{s,t}$-minor free graphs." pith.science (2026). https://pith.science/paper/AI4IPHAI
@misc{pith2026190804221,
author = {Pith},
title = {Pith review of: On the signless Laplacian spectral radius of $K_s,t$-minor free graphs},
year = {2026},
howpublished = {\url{https://pith.science/paper/AI4IPHAI}},
note = {Machine review of arXiv:1908.04221}
}
abstract
In this paper, we prove that if $G$ is a $K_{2,t}$-minor free graph of order $n\geq t^2+4t+1$ with $t\geq 3$, the signless Laplacian spectral radius $q(G)\leq \frac{1}{2}(n+2t-2+\sqrt{(n-2t+2)^2+8t-8}\ )$ with equality if and only if $n\equiv 1~(\mathrm{mod}~t)$ and $G=F_{2,t}(n)$, where $F_{s,t}(n):=K_{s-1}\vee (p\cdot K_t\cup K_r)$ for $n-s+1=pt+r$ and $0\leq r<t$. In particular, if $t=3$ and $n\geq 22$, then $F_{2,3}(n)$ is the unique $K_{2,3}$-minor free graph of order $n$ with the maximum signless Laplacian spectral radius. In addition, $F_{3,3}(n)$ is the unique extremal graph with the maximum signless Laplacian spectral radius among all $K_{3,3}$-minor free graphs of order $n\ge 1186$.
Reference graph
Works this paper leans on
-
[16]
G. Yu, J. Wang, S.-G. Guo, Maxima of the signless Laplacian spect ral radius for planar graphs, Electron. J. Linear Algebra 30 (2015) 795–811. 11
work page 2015
-
[1]
V. I. Benediktovich, Spectral radius of K2,4-minor free graph, Dokl. Nats. Akad. Nauk Belarusi 59 (2015) 5–12 (in Russian)
work page 2015
- [2]
-
[3]
J. A. Bondy, U. S. R. Murty, Graph Theory, Springer, New York , 2007
work page 2007
-
[4]
M. Chudnovsky, B. Reed, P. Seymour, The edge-density for K2,t minors, J. Combin. Theory Ser. B 101 (2011) 18–46
work page 2011
-
[5]
G. Ding, T. Johnson, P. Seymour, Spanning trees with many leave s, J. Graph Theory 37 (2001) 189–197. 10
work page 2001
-
[6]
Fang, Bounds of eigenvalues of K3,3-minor free graphs, J
K.-F. Fang, Bounds of eigenvalues of K3,3-minor free graphs, J. Inequal. Appl. 2009 (2009) 1–6
work page 2009
-
[7]
M. A. A. de Freitas, V. Nikiforov, L. Patuzzi, Maxima of the Q-index: forbidden 4-cycle and 5-cycle, Electron. J. Linear Algebra 26 (2013) 905–916
work page 2013
Show all 16 references
-
[8]
M. A. A. de Freitas, V. Nikiforov, L. Patuzzi, Maxima of the Q-index: graphs with no Ks,t, Linear Algebra Appl. 496 (2016) 381–391
2016
-
[9]
Hong, Tree-width, clique-minors, and eigenvalues, Discrete M ath
Y. Hong, Tree-width, clique-minors, and eigenvalues, Discrete M ath. 274 (2004) 281– 287
2004
-
[10]
Merris, A note on Laplacian graph eigenvalues, Linear Algebra Appl
R. Merris, A note on Laplacian graph eigenvalues, Linear Algebra Appl. 285 (1998) 33–35
1998
-
[11]
Nikiforov, The spectral radius of graphs with no K2,t-minor, Linear Algebra Appl
V. Nikiforov, The spectral radius of graphs with no K2,t-minor, Linear Algebra Appl. 531 (2017) 510–515
2017
-
[12]
J. Shu, Y. Hong, The spectral radius of K4-minor free graph, Acta Math. Appl. Sinica 5 (2001) 167–175
2001
-
[13]
Tait, The Colin de Verdi` ere parameter, excluded minors, an d the spectral radius, J
M. Tait, The Colin de Verdi` ere parameter, excluded minors, an d the spectral radius, J. Combin. Theory Ser. A 166 (2019), 42-58
2019
-
[14]
Wagner, ¨Uber eine Eigenschaft der ebenen Komplexe, Math
K. Wagner, ¨Uber eine Eigenschaft der ebenen Komplexe, Math. Ann. 114 (1937 ) 570– 590
1937
-
[15]
G. Yu, J. Shu, Y. Hong, Bounds of spectral radii of K2,3-minor free graphs, Electron. J. Linear Algebra 23 (2012) 171–179
2012
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.