REVIEW 1 major objections 4 minor 52 references
Graph theoretic properties of Speyer's matroid polynomial $g_M(t)$
T0 review · 1 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read Speyer's matroid polynomial has a graph-theoretic first derivative: at -1 it counts connected components for graphic and cographic matroids.
desk verdict Solid paper with a real typesetting error in the central expansion that must be fixed, but the math holds up and the algorithmic/data contributions are strong. 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 engine of the proof is Speyer's polynomial as a covaluative matroid invariant: it is determined by its values on series-parallel matroids (all equal to $t$) and by decomposition of a matroid polytope into Schubert matroid polytopes indexed by chains in the lattice of cyclic flats. The paper's new formula (Proposition 3.6) writes the chain-lattice Möbius coefficients as products of interval Möbius values in the cyclic-flat lattice, replacing an enormous chain sum by a much smaller lattice computation. The graph-theoretic half uses Crapo's $\beta$ invariant $\beta(M)$ and Elser's identity for nuclei, an alternating-sum identity over connected vertex-covering subgraphs, to turn sums over all edge subsets into sums over blocks. This combination yields the closed expression $g'_M(-1)=(-1)^{c(M)-1}\sum_{A\subseteq M}(-1)^{\ell(A)}\beta(A)\operatorname{rk}(A)$, and Theorem 1.1 identifies the same sum with $c(M)$ for graphic and cographic matroids.
What would settle it
Evaluate $g'_M(-1)$ for any biconnected graph's cycle matroid: if it is not $1$, Theorem 1.8 is false. A sharper test would be a graphic or cographic matroid with $c(M)=2$ whose first derivative differs from $(-1)^{1}\cdot 2=-2$; likewise, to test the conjectures, search for a 3-connected planar graph with $N_2\neq 1$, or a connected cubic graph with $g''_G(0)\neq 2n\,t_{0,1}(G)-4t_{0,2}(G)$.
Extended reading notes
Core claim
On the paper's own terms, the central result is Theorem 1.8: for any graphic or cographic matroid $M$ with no loops or coloops, $g'_M(-1)=(-1)^{c(M)-1}c(M)$. Equivalently, writing $g_M(t)=t^{\operatorname{rk}(M)-1}\sum_i N_i(M)(1+t)^i$, every biconnected graph's cycle matroid has $N_0=1$ and $N_1=0$. The paper also proves a $k$-connected refinement (Theorem 2.9): for a $k$-connected graph $G$, the alternating sum over edge subsets with weights $\binom{\operatorname{rk}(A)}{k}$ counts $(k-1)$-vertex cuts. Since the first derivative is trivial for biconnected graphs, the paper identifies $N_2(G)$ as a genuinely new numerical invariant of graphs, supports it with computations over more than three million graphs, and formulates several data-driven conjectures about its behavior under planarity, 3-sums, 4-edge and 4-vertex twists, and, for cubic graphs, its relation to the flow polynomial.
Load-bearing premise
The proof of the main identity depends on the theorem that Speyer's polynomial is covaluative and can be decomposed, via the lattice of cyclic flats, into Schubert matroids; if that decomposition gave a different value for even one matroid without loops or coloops, the component-counting formula would not follow.
Editorial extensions
If this is right
- Every biconnected graph's cycle matroid satisfies $g'_M(-1)=1$ and $N_1(M)=0$, so the coefficient $N_2(M)$ becomes the first nontrivial invariant in the expansion about $t=-1$.
- A connected matroid with $g'_M(-1)\neq 1$, such as the uniform matroid $U^n_r$ with $2\leq r\leq n-2$, cannot be graphic or cographic.
- Theorem 2.9 gives a $k$-connected refinement: for a $k$-connected graph, the corresponding alternating sum is $1$ for all smaller connectivity levels and jumps only when $k$-vertex cuts are deleted, counting those cuts weighted by their number of extra components.
- The improved algorithm reduces the computation of $g_M(t)$ to cyclic-flat lattice data rather than all chains of cyclic flats; graphs such as the wheel $W_{18}$, whose cyclic-flat chain lattice has roughly $1.77\times 10^{13}$ elements, become computable in under two hours, and the resulting open data set covers over three million small graphs.
- For 1- and 2-sums, $N_2$ decomposes additively (after a correction term), so $N_2$ of any graph is determined by its 3-connected components together with the component count.
Reading between the lines
- If Theorem 1.8 is correct, the single evaluation $g'_M(-1)$ gives a cheap necessary condition for a matroid to be graphic or cographic, and the $N_1=0$ condition for biconnected graphs could serve as a fast filter in algorithms that search for graphic or cographic representations.
- The conjectured reduction rules for $N_2$ under 3-sums, twists, and vertex deletions mirror exactly the known identities for Feynman period integrals; if the conjectures survive, $N_2$ would be a purely combinatorial invariant in the same relation web as the $c_2$-invariant, the Martin sequence, and the Hepp bound, and one could test whether $N_2$ is determined by any of those invariants.
- The cubic-graph relation between $g''_G(0)$ and the flow polynomial, if true, suggests that low-degree Tutte-polynomial relations may reappear in other restricted graph classes (for example 4-regular graphs) even though no linear relation exists for all biconnected graphs.
- The observation that $N_2(G)=1$ for all tested 3-connected planar graphs while $K_5$ and $K_{3,3}$ have $N_2=0$ raises the possibility that $N_2$, together with the 3-sum and twist rules, could yield an algebraic invariant that distinguishes some non-planar graphs from planar ones in a valuative, polytope-compatible way.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies Speyer's matroid polynomial g_M(t) for graphic and cographic matroids. It proves that Crapo's beta invariant satisfies the graph-theoretic identity beta(M) rk(M) = (-1)^{rk(M)} sum_A (-1)^{|A|} c(A) (Theorems 1.1 and 2.5), generalizes this to a higher-connectivity identity involving weighted counts of vertex cuts (Theorem 1.2 / 2.9), and derives a formula for g'_M(-1) in terms of an alternating sum of beta and rank values (Proposition 1.5). The combination yields g'_M(-1) = (-1)^{c(M)-1} c(M) for every graphic or cographic matroid without loops or coloops (Theorem 1.8). The paper also proposes a substantially improved recursive algorithm for computing g_M(t), provides an open-source implementation and a public dataset of more than three million graphs, and uses the resulting data to formulate several explicit conjectures about the coefficient N_2(G), including a planarity prediction, 3-sum and twist identities, a relation with the flow polynomial for cubic graphs, and connections to Feynman integrals.
Significance. If the results hold, Theorem 1.8 provides one of the few exact evaluations of Speyer's matroid polynomial and yields a simple algebraic obstruction to graphic or cographic representability. The improved algorithm is a genuine practical advance: the paper demonstrates that the wheel graph W_18, whose lattice of cyclic flats has about 17.7 trillion chains, can be handled in under two hours, whereas the previous chain-sum algorithm is impractical. The conjectures identify N_2(G) as a rich graph invariant that interacts with planarity, vertex and edge cuts, flow polynomials, and Feynman periods. The paper ships open-source Maple code and a public dataset, and the implementation is cross-checked against an independent Sage program on 81 graphs and against Tutte-polynomial computations of beta on the entire dataset of more than three million graphs. The conjectures are explicitly labeled as conjectures, are supported by substantial computations, and are falsifiable.
major comments (1)
- The displayed expansion g_M(t) = t^{rk(M)-1} sum_i N_i(M)(1+t)^i is inconsistent with the rest of the paper. For the cycle matroid of K_4, the paper itself computes g_{K_4}(t)=2t+2t^2+t^3 in Example 3.13; this polynomial has a nonzero t-coefficient, while the right-hand side with t^{rk(M)-1}=t^2 would be divisible by t^2. The same contradiction occurs for the rank-3 series-parallel matroid with g(t)=t. All subsequent uses, including Corollaries 3.14 and 3.15, Lemma 3.17, Example 3.16, and the identity N_0(M)=(-1)^{c(M)-1}, are consistent only with the corrected expansion g_M(t)=t sum_i N_i(M)(1+t)^i. Since the proof of Proposition 1.5 and hence Theorem 1.8 is expressed through equation (3.8), the proof is not formally checkable as typeset. The correction appears to be purely notational, but it must be made consistently in equations (1.5) and (3.8) and in the surrounding discussion.
minor comments (4)
- The formulas for the prisms and Möbius ladders appear to have missing superscripts: '2n-n-3' and '2n-n-1' should likely read '2^n-n-3' and '2^n-n-1', and the subsequent beta values '2n-n-1' and '2n-n' should likely read '2^n-n-1' and '2^n-n'. As written, the displayed polynomials reduce to expressions involving n-3 and n-1, which are inconsistent with the stated beta values.
- There are several typographical errors that should be corrected: 'irreducbile' in Remark 1.17, 'Corollay 1' in the citation for Lemma 3.18, 'ennumerating' in Section 3.4, 'Delanny' in Corollary 3.14, and 'illsutrated' in Section 4.2.
- The column headers of Table 4 are difficult to parse because adjacent columns are both labeled 'G' and 'N2 G'; please restructure the table or clarify the convention so that each vertex count is unambiguously paired with its N_2 value.
- The argument that the star-triangle identity implies Conjecture 1.13 relies on the Steinitz reduction sequence; it would be helpful to make explicit that each move in the cited sequence preserves the condition |pi0(G_i \ S)| = 2, which is the condition needed to apply equation (1.7).
Circularity Check
No circularity found: the main theorems are derived from external results and independent graph-theoretic identities.
full rationale
The central claim, Theorem 1.8, is obtained by combining two genuinely independent derivations. Proposition 1.5 expresses g'_M(-1) as (-1)^{c(M)-1} times an alternating sum over subsets; its proof uses the covaluative characterization of Speyer's polynomial from [47], [21], and [19], which are external works, not self-citations, and the expansion coefficients are computed from the lattice of cyclic flats rather than fitted to the target identity. The second ingredient, Theorem 1.1 (via Theorem 2.5 and Theorem 2.6), evaluates the same alternating sum as c(M) for graphic or cographic matroids using Elser's identity, Crapo's beta invariant, and inclusion-exclusion; this part makes no reference to g_M(t), so there is no definitional loop. The conjectures in Section 4 are explicitly empirical statements supported by large data sets and are not presented as forced predictions; they therefore do not constitute circularity. The only notable issue is a typesetting inconsistency in the displayed expansion (1.5) and (3.8), where the prefactor t^{rk(M)-1} conflicts with the examples (e.g., g_{K4}(t)=2t+2t^2+t^3) and with the later corrected usage g_M(t)=t sum N_i(M)(1+t)^i. This is a correctness or typographical concern for the proof as written, not a case of the result being equivalent to its inputs by construction, and it does not affect the circularity score.
Assumptions & free parameters
assumptions (3)
- domain assumption Speyer's polynomial g_M(t) is a well-defined covaluative matroid invariant satisfying the three axioms in Section 3 (loop/coloop vanishing, series-parallel value t, multiplicativity over direct sums).
- domain assumption The decomposition of a matroid into Schubert matroids via its lattice of cyclic flats (Theorem 3.5 from [19]) is valid and gives an explicit evaluation of g_M.
- standard math Standard results from matroid theory and graph theory, including Crapo's beta invariant, matroid duality, the block-cut tree formula, and Möbius inversion in lattices.
Cite this review
Pith. "Pith review of Graph theoretic properties of Speyer's matroid polynomial $g_M(t)$." pith.science (2026). https://pith.science/paper/KEQE7AYP
@misc{pith2026250618788,
author = {Pith},
title = {Pith review of: Graph theoretic properties of Speyer's matroid polynomial $g_M(t)$},
year = {2026},
howpublished = {\url{https://pith.science/paper/KEQE7AYP}},
note = {Machine review of arXiv:2506.18788}
}
abstract
We prove relations between the number of $k$-connected components of a graph, Crapo's invariant $\beta(M)$ of a matroid, and Speyer's polynomial $g_M(t)$. These yield a simple interpretation of $g_M'(-1)$ when $M$ is graphic or cographic. Furthermore, we improve Ferroni's algorithm to compute $g_M(t)$ and provide an implementation and an extensive data set. These calculations reveal a large number of graph theoretic constraints on the second derivative $g_M''(-1)$, which we thus advertise as an intriguing new invariant of graphs. We also propose a relation between the flow polynomial and $g_M''(0)$ for cubic graphs.
Figures
Figures from the paper (13 more)
Reference graph
Works this paper leans on
-
[1]
The external activity complex of a pair of matroids
A. Berget and A. Fink, “The external activity complex of a pair of matroids.” preprint, Dec., 2024, arXiv:2412.11759 [math.CO]
arXiv 2024
-
[2]
N. L. Biggs, R. M. Damerell and D. A. Sands,Recursive families of graphs, J. Comb. Theory, Ser. B12 (1972), no. 2 pp. 123–131
work page 1972
- [3]
-
[4]
J. E. Bonin and A. de Mier,Lattice path matroids: Structural properties, Eur. J. Comb. 27 (2006), no. 5 pp. 701–738, arXiv:math/0403337 [math.CO]
work page Pith review arXiv 2006
-
[5]
J. E. Bonin and A. de Mier,The lattice of cyclic flats of a matroid, Annals of Combinatorics 12 (July, 2008) pp. 155–170, arXiv:math/0505689 [math.CO]
arXiv 2008
-
[6]
M. Borinsky and O. Schnetz,Recursive computation of Feynman periods, JHEP 2022 (Aug., 2022) p. 291, arXiv:2206.10460 [hep-th]
arXiv 2022
-
[7]
G.Brinkmannand B.D.McKay, Fast generation of planar graphs, MATCH Commun. Math. Comput. Chem.58 (2007), no. 2 pp. 323–357. Program available athttp: //cs.anu.edu.au/~bdm/plantri
work page 2007
-
[8]
T. Brylawski and J. Oxley,The Tutte polynomial and its applications, inMatroid applications, Encyclopedia of Mathematics and its Applications, pp. 123–225. Cam- bridge University Press, 1992
work page 1992
Show all 52 references
-
[9]
Chopra,On the spanning tree polyhedron, Operations Research Letters8 (1989), no
S. Chopra,On the spanning tree polyhedron, Operations Research Letters8 (1989), no. 1 pp. 25–29
1989
-
[10]
Coolsaet, S
K. Coolsaet, S. D’hondt and J. Goedgebeur,House of Graphs 2.0: A database of interesting graphs and more, Discrete Applied Mathematics325 (2023) pp. 97–107. Available athttps://houseofgraphs.org/
2023
-
[11]
H. H. Crapo,A higher invariant for matroids, Journal of Combinatorial Theory2 (1967), no. 4 pp. 406–417
1967
-
[12]
H. H. Crapo,Möbius inversion in lattices, Arch. Math.19 (1969) pp. 595–607
1969
-
[13]
Crump,Properties of the extended graph permanent, Commun
I. Crump,Properties of the extended graph permanent, Commun. Num. Theor. Phys. 11 (2017), no. 4 pp. 791–836, arXiv:1608.01414 [math.CO]
2017 arXiv
-
[14]
Crump, M
I. Crump, M. DeVos and K. Yeats,Period preserving properties of an invariant from the permanent of signed incidence matrices, Ann. Inst. H. Poincaré D3 (2016), no. 4 pp. 429–454, arXiv:1505.06987 [math.CO]
2016 arXiv
-
[15]
W. H. Cunningham and J. Edmonds,A combinatorial decomposition theory, Cana- dian J. Math.32 (June, 1980) pp. 734–765
1980
-
[16]
Dorpalen-Barry, C
G. Dorpalen-Barry, C. Hettle, D. C. Livingston, J. L. Martin, G. D. Nasr, J. Vega and H. Whitlatch,A positivity phenomenon in Elser’s Gaussian-cluster percolation model, J. Comb. Theory, Ser. A179 (2021) p. 105364, arXiv:1905.11330 [math.CO]
2021 arXiv
-
[17]
J. N. Eberhardt,Computing the Tutte polynomial of a matroid from its lattice of cyclic flats, Electron. J. Comb.21 (2014), no. 3 p. P3.47, arXiv:1407.6666 [math.CO]
2014 arXiv
-
[18]
Elser,Gaussian-cluster models of percolation and self-avoiding walks, Journal of Physics A: Mathematical and General17 (May, 1984) pp
V. Elser,Gaussian-cluster models of percolation and self-avoiding walks, Journal of Physics A: Mathematical and General17 (May, 1984) pp. 1515–1523. 53
1984
-
[19]
Ferroni,Schubert matroids, Delannoy paths, and Speyer’s invariant, Comb
L. Ferroni,Schubert matroids, Delannoy paths, and Speyer’s invariant, Comb. Theory 3 (2023), no. 3 p. 13, arXiv:2311.01397 [math.CO]
2023 arXiv
-
[20]
The polytope of all matroids
L. Ferroni and A. Fink, “The polytope of all matroids.” preprint, Feb., 2025, arXiv:2502.20157 [math.CO]
2025 arXiv
-
[21]
Ferroni and B
L. Ferroni and B. Schröter,Valuative invariants for large classes of matroids, J. Lond. Math. Soc., II. Ser.110 (2024), no. 3 p. e12984, arXiv:2208.04893 [math.CO]
2024 arXiv
-
[22]
The omega invariant of a matroid
A. Fink, K. Shaw and D. E. Speyer, “The omega invariant of a matroid.” preprint, Nov., 2024, arXiv:2411.19521 [math.CO]
2024
-
[23]
Fink and D
A. Fink and D. E. Speyer,K-classes for matroids and equivariant localization, Duke Math. J.161 (Nov., 2012) pp. 2699–2723, arXiv:1004.2403 [math.CO]
2012 arXiv
-
[24]
Freij-Hollanti, M
R. Freij-Hollanti, M. Grezet, C. Hollanti and T. Westerbäck,Cyclic flats of binary matroids, Adv. Appl. Math.127 (2021) p. 102165, arXiv:1906.10936 [math.CO]
2021 arXiv
-
[25]
Georgiadis, K
L. Georgiadis, K. Giannis, G. F. Italiano and E. Kosinas,Computing vertex-edge cut- pairs and 2-edge cuts in practice, in19th International Symposium on Experimental Algorithms (SEA 2021) (D. Coudert and E. Natale, eds.), vol. 190 of Leibniz International Proceedings in Inform...
2021
-
[26]
Grinberg,The Elser nuclei sum revisited, Discrete Math
D. Grinberg,The Elser nuclei sum revisited, Discrete Math. Theor. Comput. Sci. 23 (2021), no. 1 p. #15, arXiv:2009.11527 [math.CO]
2021 arXiv
-
[27]
Grünbaum, Convex polytopes, vol
B. Grünbaum, Convex polytopes, vol. Vol. 16 ofPure and Applied Mathematics. Interscience Publishers John Wiley & Sons, Inc., New York, 1967. With the cooperation of Victor Klee, M. A. Perles and G. C. Shephard
1967
-
[28]
Guo and B
K. Guo and B. Mohar,Large regular bipartite graphs with median eigenvalue 1, Linear Algebra Appl.449 (2014) pp. 68–75, arXiv:1309.7025 [math.CO]
2014 arXiv
-
[29]
Hampe,The intersection ring of matroids, J
S. Hampe,The intersection ring of matroids, J. Comb. Theory, Ser. B122 (2017) pp. 578–614, arXiv:1602.07167 [math.CO]
2017 arXiv
-
[30]
Harary,An elementary theorem on graphs, Am
F. Harary,An elementary theorem on graphs, Am. Math. Mon.66 (1959), no. 5 pp. 405–407
1959
-
[31]
Jurrius,Relations between Möbius and coboundary polynomials, Math
R. Jurrius,Relations between Möbius and coboundary polynomials, Math. Comput. Sci. 6 (2012), no. 2 pp. 109–120, arXiv:1202.3303 [math.CO]
2012 arXiv
-
[32]
Kochol, Decomposition formulas for the flow polynomial, Eur
M. Kochol, Decomposition formulas for the flow polynomial, Eur. J. Comb.26 (2005), no. 7 pp. 1086–1093
2005
-
[33]
J. P. S. Kung, G.-C. Rota and C. H. Yan,Combinatorics: The Rota Way. Cambridge Mathematical Library. Cambridge University Press, 2009. 54
2009
-
[34]
B. D. McKay and A. Piperno,Practical graph isomorphism, II, J. Symb. Comput. 60 (Jan., 2014) pp. 94–112, arXiv:1301.1493 [cs.DM]. Program website: http: //pallini.di.uniroma1.it/
2014 arXiv
-
[35]
Merino, A
C. Merino, A. de Mier and M. Noy,Irreducibility of the Tutte polynomial of a connected matroid, J. Comb. Theory, Ser. B83 (2001), no. 2 pp. 298–304
2001
-
[36]
Pub- lished electronically athttp://oeis.org
OEIS Foundation Inc.,The On-Line Encyclopedia of Integer Sequences, 2025. Pub- lished electronically athttp://oeis.org
2025
-
[37]
J. G. Oxley,Matroid theory, vol. 21 ofOxford Graduate Texts in Mathematics. Oxford University Press, 2nd ed., 2011
2011
-
[38]
Panzer,Hepp’s bound for Feynman graphs and matroids, Ann
E. Panzer,Hepp’s bound for Feynman graphs and matroids, Ann. Inst. H. Poincaré D 10 (2023), no. 1 pp. 31–119, arXiv:1908.09820 [math-ph]
2023 arXiv
-
[39]
Panzer, Speyer polynomials of graphs, 2025
E. Panzer, Speyer polynomials of graphs, 2025. data set, available at https:// people.maths.ox.ac.uk/panzer/data/g.tar.gz
2025
-
[40]
Panzer and O
E. Panzer and O. Schnetz,The Galois coaction onϕ4 periods, Commun. Num. Theor. Phys.11 (2017), no. 3 pp. 657–705, arXiv:1603.04289 [hep-th]
2017 arXiv
-
[41]
Panzer and K
E. Panzer and K. Yeats,Feynman symmetries of the Martin andc2 invariants of regular graphs, Combinatorial Theory 5 (2025), no. 1 p. #10, arXiv:2304.05299 [math.CO]
2025 arXiv
-
[42]
Schnetz, Quantum periods: A Census ofϕ4-transcendentals, Commun
O. Schnetz, Quantum periods: A Census ofϕ4-transcendentals, Commun. Num. Theor. Phys.4 (2010), no. 1 pp. 1–47, arXiv:0801.2856 [hep-th]
2010 arXiv
-
[43]
Schnetz, Quantum field theory overFq, Electron
O. Schnetz, Quantum field theory overFq, Electron. J. Combin.18 (May, 2011) p. P102, arXiv:0909.0905 [math.CO]
2011 arXiv
-
[44]
Sekine and C
K. Sekine and C. Q. Zhang,Decomposition of the flow polynomial, Graphs and Combinatorics 13 (June, 1997) pp. 189–196
1997
-
[45]
P. D. Seymour,Decomposition of regular matroids, J. Comb. Theory, Ser. B28 (June, 1980) pp. 305–359
1980
-
[46]
L. S. Shapley,Cores of convex games, Int. J. Game Theory1 (1971) pp. 11–26
1971
-
[47]
D. E. Speyer,A matroid invariant via theK-theory of the Grassmannian, Adv. Math. 221 (2009), no. 3 pp. 882–913, arXiv:math/0603551 [math.AG]
2009 arXiv
-
[48]
R. P. Stanley,Enumerative Combinatorics, Volume 1. Cambridge Studies in Ad- vanced Mathematics. Cambridge University Press, 2 ed., 2011
2011
-
[49]
J. R. Swenson,The chromatic polynomial of a complete bipartite graph, Am. Math. Mon. 80 (1973), no. 7 pp. 797–798. 55
1973
-
[50]
http://www.sagemath.org/
The Sage Developers, SageMath, the Sage Mathematics Software System, 2024. http://www.sagemath.org/
2024
-
[51]
Chain Tutte polynomials
M. Wakefield, “Chain Tutte polynomials.” preprint, 2023, arXiv:2305.02874 [math.CO]
2023 arXiv
-
[52]
E. G. Whitehead,Chromatic polynomials for chorded cycles, inProceedings of the Sixth Southeastern Conference on Combinatorics, Graph Theory and Computing (Florida Atlantic Univ., Boca Raton, Fla., 1975), vol. XIV ofCongress. Numer., pp. 619–625, Utilitas Math., Winnipeg, MB, 1975. 56
1975
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.