Hilbert Series and Logarithmic Degrees of $A$-Hypergeometric Series

Mao Nagamine

REVIEW 2 major objections 4 minor 10 references

Hilbert Series and Logarithmic Degrees of $A$-Hypergeometric Series

T0 review · 2 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read For every fake exponent, the graded dimensions of the local solution space are the coefficients of a Hilbert series of an Artinian Stanley–Reisner quotient.

desk verdict A genuinely useful formal theorem: the Hilbert series of the local fake indicial ideal is computed by an Artinian Stanley–Reisner quotient with no Cohen–Macaulay hypothesis, though the advertised bridge to actual logarithmic series rests on an imported condition that is only sketched. read the letter →

arxiv 2608.01778 v1 pith:DVF5CBVQ submitted 2026-08-03 math.AC math.AG

Mao Nagamine This is my paper · ORCID

classification math.ACmath.AG MSC 33C7013F5513H1014M25

keywords A-hypergeometricsystemlogarithmicseriesfakeexponentstandardpairStanley-ReisnerringHilbertOkuyama-SaitoFrobeniusconditionCohen-Macaulay

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

This paper proves that the logarithmic content of an A-hypergeometric system near any fake exponent is governed by one polynomial: the Hilbert series of an Artinian quotient built from the Stanley–Reisner ring of the link of the exponent's negative support. The link is extracted from all standard pairs of the initial toric ideal that correspond to that exponent, including embedded standard pairs. The identity H_{w,v}(t)=sum_q dim(Q_v^perp)_q t^q holds without any Cohen–Macaulay hypothesis, and under the paper's Frobenius condition the same series gives the graded dimensions of the actual leading logarithmic coefficient space. The result matters because it converts a hard analytic Frobenius computation into a finite quotient computation, and it identifies exactly when that computation collapses to the familiar h-polynomial.

What carries the argument

The carrier of the argument is the Artinian Stanley–Reisner quotient A_{w,v}, defined through the simplicial complex link_v: the link is obtained by deleting the negative-support variables I0 from the supports of standard pairs in S_w(v). The linear forms ell_i come from row-reducing the Euler relations so that the I0-variables are eliminated, and the identity M_theta(v)=I_{link_v} identifies the residual monomial relations with the Stanley–Reisner ideal. The proof of Artinianity uses the linear independence of the columns of A over I0 union tau for every face tau, so the forms cut out the origin on the complex. The perfect pairing between T and U then converts the Hilbert series of the quot

What would settle it

For a concrete check of the main identity, take any fake exponent v with an explicitly known Grobner basis, compute the quotient C[link_v]/(ell_i) by row-reducing the Euler relations, and compare its Hilbert series with the graded dimensions of Q_v^perp from the generators of the monomial and Euler ideals; Theorem 4.7 says they agree degree by degree, so a single disagreement at any q would settle that the claim is false.

Watch

Extended reading notes

Core claim

For any fake exponent v, the paper builds an Artinian quotient A_{w,v}=C[link_v]/(ell_1,...,ell_{d-|I0|}) from all standard pairs that match v, including embedded ones. Its Hilbert series H_{w,v}(t) is shown to equal sum_q dim(Q_v^perp)_q t^q, the graded dimensions of the orthogonal complement of the local fake indicial ideal. Under the Frobenius condition P=m(s)P_B, that same polynomial gives the graded dimensions of the leading logarithmic coefficient space C_v of actual series solutions. If a top-dimensional standard pair occurs and C[link_v] is Cohen–Macaulay, H_{w,v}(t) is the h-polynomial of link_v.

Load-bearing premise

Everything about actual series solutions rests on condition (2.1), P=m(s)P_B; the paper assumes it when needed and does not say when it can be expected to hold.

Editorial extensions

If this is right

For any fake exponent, the graded dimensions dim(Q_v^perp)_q are the coefficients of a polynomial Hilbert series H_{w,v}(t), computed by a finite quotient of a Stanley–Reisner ring.
When the Frobenius condition (2.1) holds, this same polynomial gives the graded dimensions of the leading logarithmic polynomial space of actual series solutions, so logarithmic degrees are read off without constructing all series.
If a fake exponent comes from a top-dimensional standard pair and the link is Cohen–Macaulay, H_{w,v}(t) equals the h-polynomial of the link, so logarithmic degrees are the degree and evaluation of that h-polynomial.
The construction is algorithmic: compute standard pairs including embedded ones, remove the negative-support variables, row-reduce the Euler relations, and compute the Hilbert series of the resulting Artinian quotient.
The theorem holds even for fake exponents supported only on embedded standard pairs, where the number of Euler forms exceeds the complex dimension and the h-polynomial formula fails.

Reading between the lines

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

The quotient construction could be used as a fast numerical probe for the Frobenius condition: if the Hilbert series computed combinatorially disagrees with a direct series computation, that specific parameter violates (2.1).
A natural open step left implicit by the paper would be a purely combinatorial test for P=m(s)P_B, perhaps read from the inclusion poset of negative supports NS_w(v), instead of from the Frobenius ideal itself.
For embedded fake exponents, the degree of H_{w,v}(t) behaves like a 'logarithmic depth' contributed by embedded components; in the examples it is lower than the dimension of the link, so it may serve as a numerical measure of how far the exponent is from being principal.

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

Dear colleague,

The real result here is Theorem 4.7, and it appears to be correct and genuinely new. For every fake exponent, the paper constructs a finite-dimensional graded quotient of the Stanley–Reisner ring of the link, built from all corresponding standard pairs, embedded ones included, and proves its Hilbert series equals that of the orthogonal complement of the local fake indicial ideal. That removes the Cohen–Macaulay assumption that limited the earlier Saito–Sturmfels–Takayama approach. The proof is honest: the Artinian property via rank arguments on the submatrix B'_tau is clean, the elimination of the negative-support variables is handled explicitly, and the non-CM example in 6.1 shows exactly where the h-polynomial specialization fails. The paper earns credit for the formal part, and the AI-use disclosure is refreshingly straightforward.

The soft spot is exactly where the reader's stress-test points. The headline claim about actual logarithmic solution spaces C_v depends entirely on the Okuyama–Saito condition (2.1), imported from [6] and only sketched in Theorem 3.5. The author does not characterize when (2.1) holds, and the cited verification in Example 6.1 relies on an unpublished companion paper [5]. If Theorem 3.5 in [6] carries hidden extra hypotheses, or if the isomorphism involves a degree shift when K is properly contained in I_0, then Corollary 4.8 does not follow from the present text. That is a genuine gap in the advertised story, but it is explicitly flagged as an assumption, and Theorem 4.7 stands independently as a statement about the formal local indicial data.

Minor issues: the notation is dense in places, and the reader has to keep track of several different P's and Q's. The examples are helpful but only partially verify the Gröbner basis claims; an expert in the specific literature would need to check those computations. The self-citation [5] is used only in an example, so the circularity burden is low.

Who is this for? Anyone working in hypergeometric systems or combinatorial commutative algebra who wants a per-exponent graded Hilbert series without CM assumptions. The formal theorem deserves a serious referee; the Frobenius bridge needs a separate careful check against [6].

My recommendation: send to peer review. The formal part is solid and novel, and the caveat about (2.1) is a condition to verify, not a hidden contradiction.

Best, [You]

Referee Report

2 major / 4 minor

Summary. The paper fixes a generic weight vector w and a fake exponent v of a homogeneous A-hypergeometric system. It forms the set S_w(v) of standard pairs of in_w(I_A) compatible with v, lets I_0 be the negative support of v, and builds the link_v of I_0 in the simplicial complex spanned by these supports. The main object is the Artinian quotient A_{w,v} = C[link_v]/(ell_1,...,ell_{d-|I_0|}), where the ell_i are the V-variable parts of the Euler relations after eliminating the I_0 variables. Theorem 4.7 asserts Hilb(A_{w,v},t) = sum_q dim(Q_v^perp)_q t^q, where Q_v is the local fake indicial ideal. Under the Okuyama-Saito equality P = m(s)P_B (2.1), Corollary 4.8 identifies the same series with the graded dimension series of the actual logarithmic coefficient space C_v. Section 5 proves that under a top-dimensional standard-pair hypothesis and Cohen-Macaulayness of link_v, the series is the h-polynomial; Section 6 presents one non-CM, one embedded-pair, and two CM examples.

Significance. The formal result, if fully justified, is a significant and elegant step: it gives a finite, purely combinatorial model for the graded lengths of the local logarithmic solution space at a fixed fake exponent, with no Cohen-Macaulay assumption and no restriction to top-dimensional standard pairs. The rank argument in Lemma 4.5 is convincing, and the examples, including the non-CM case where the Hilbert series is not the h-polynomial, are instructive. The construction is explicit and algorithmic (Algorithm 5.5). The paper is also honest about the external Frobenius condition, which is the main limitation: without (2.1), only Q_v^perp is computed. The principal weakness is the proof of Theorem 3.5, which is a sketch of an imported result but is load-bearing for the title claim.

major comments (2)

[Section 3.3, Theorem 3.5] This theorem is the only bridge from the formal quotient Q_v^perp to the actual coefficient space C_v, and hence the basis for Corollary 4.8 and the abstract's statement about logarithmic degrees. The proof is a citation sketch: it does not demonstrate why, under P = m(s)P_B, the operators contributing to the starting monomial are exactly P_B^perp rather than P^perp or some quotient of it, nor why this identification is degree-preserving when m(s) has positive degree (K strictly contained in I_0). Since (2.1) allows m(s) != 1, the degree-shift issue is real and needs an explicit argument or a precise reference to Okuyama-Saito [6] that states exactly this degree-preserving isomorphism. Please either prove Theorem 3.5 or quote it verbatim from [6] with all hypotheses.
[Section 4.1, Lemma 4.1] The proof concludes 'a homogeneous ideal is recovered from its localization at the homogeneous maximal ideal.' This statement is false for arbitrary homogeneous ideals. In the present situation both ideals are monomial in theta, and for monomial ideals the conclusion does follow: if a monomial x^a is a minimal generator of one ideal and the localizations agree, then a unit times x^a lies in the other, forcing x^a to be a multiple of a minimal generator. Please replace the incorrect general claim by the monomial-ideal argument. Since Lemma 4.1 feeds directly into Theorem 4.2(ii) and Proposition 4.4, the proof should be repaired.

minor comments (4)

[Section 2.3, Definition of NS_w(v)] The definition is hard to parse because the quantifiers over u and u' are not fully explicit. Please rephrase, for example: 'I_u belongs to NS_w(v) if every u' in L with I_{u'} = I_u lies in C(w).'
[Section 6.3, Reference [5]] The paper uses the unpublished reference [5] to conclude K = I_0 in Example 6.3. If [5] is not yet available, the example should be self-contained or the conclusion should be marked as conditional on [5].
[Section 6.2, Example 6.2] The sentence 'The two Euler forms do not form a regular sequence: after quotienting by the first, the image of the second is zero' is correct but terse. Please clarify that the image is zero, hence not a nonzero-divisor, in the quotient ring.
[Section 4.1, Lemma 4.1 final sentence] The phrase 'a homogeneous ideal is recovered from its localization' should be changed to 'a monomial ideal is recovered from its localization' to make the argument valid; see the major comment.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the Hilbert-series theorem is derived from standard pairs, Euler-relation elimination, and the monomial pairing; the upgrade to actual logarithmic solutions is explicitly conditional on the external Okuyama–Saito hypothesis (2.1).

full rationale

The central formal result, Theorem 4.7, is self-contained and does not reduce to its conclusion. Q_v is defined in Section 3.3 as ⟨Aθ⟩ + M_θ(v). Lemma 4.1 identifies M_θ(v) with the intersection of the localized prime components ⟨θ_j | j∉σ⟩ over S_w(v); Theorem 4.2(ii) identifies that intersection with the Stanley–Reisner ideal of link_v; Lemma 4.3 and Proposition 4.4 eliminate the I_0-variables to produce an isomorphism T/Q_v ≅ C[link_v]/(ℓ_1,...,ℓ_{d−|I0|}); and Proposition 3.4 gives dim(Q_v^⊥)_q = dim(T/Q_v)_q from the perfect monomial pairing. No parameter is fitted to the target Hilbert series, and the linear forms ℓ_i come from row-reducing the Euler relations, not from the claimed dimensions. Lemma 4.5 independently proves the quotient is Artinian. Corollary 4.8 upgrades Q_v^⊥ to the actual space C_v only under the imported Okuyama–Saito condition P = m(s)P_B in (2.1). That condition is an external hypothesis from [6], used as a sufficient assumption rather than as an output of the construction, so this is a conditional dependence and a possible correctness risk if (2.1) fails, not circularity. The only self-citation, [5], appears in Example 6.3 to verify K = I_0 in one illustrative example; it is not load-bearing for the main theorem. No circular step can be exhibited by equation reduction or by definitional identification, so the paper receives a low score for a minor self-citation rather than for actual circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The construction takes A, w, v as input; no constants are fitted. The central Theorem 4.7 relies on standard-pair decomposition of the distraction and on localizing at the homogeneous maximal ideal, both from [9] and [4]. The actual-logarithmic corollary additionally assumes the Okuyama-Saito equality (2.1) and the imported theorem from [6]. One self-cited to-appear reference [5] is used only in Example 6.3.

assumptions (6)

domain assumption A has rank d and its columns lie in an affine hyperplane of Q^d not containing the origin (Section 1).
Defines the homogeneous A-hypergeometric system and makes the D-module regular holonomic (cited [3,10]).
domain assumption w is a generic weight vector and v is a fake exponent of HA(beta), so in_w(I_A) is the reduced Groebner initial ideal and Av = beta (Sections 2.1-2.2).
All standard pairs S_w(v) and the link are built from these data.
standard math Standard pair decomposition of the distraction: fM = intersection over (a,sigma) in S(M) of <theta_j - a_j | j not in sigma> [9, Cor 3.2.3].
Load-bearing for the identity J_w(v) = I_link_v in Lemma 4.1.
standard math A homogeneous ideal in a positively graded C-algebra is recovered from its localization at the homogeneous maximal ideal.
Bridges the localized identification in Lemma 4.1 to the global equality M_theta(v) = intersection of <theta_j | j not in sigma>.
standard math Reisner criterion, shellability, and CM regularity facts for h-polynomials (Proposition 3.3, equation (3.1), [4]).
Used in Section 5 and in the examples to obtain h-polynomial specializations.
domain assumption Okuyama-Saito Theorem 3.11 and Theorem 4.4 of [6]: if P = m(s)P_B then all series solutions with exponent v are obtained by the Frobenius calculus and C_v is graded-isomorphic to Q_v^perp.
This is the only bridge from formal fake indicial data to actual series solutions; it is assumed rather than reproved (Theorem 3.5).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Hilbert Series and Logarithmic Degrees of $A$-Hypergeometric Series." pith.science (2026). https://pith.science/paper/DVF5CBVQ

@misc{pith2026260801778,
  author       = {Pith},
  title        = {Pith review of: Hilbert Series and Logarithmic Degrees of $A$-Hypergeometric Series},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DVF5CBVQ}},
  note         = {Machine review of arXiv:2608.01778}
}

abstract

Fix a generic weight vector and a fake exponent of a homogeneous $A$-hypergeometric system. Using all corresponding standard pairs, including embedded ones, we construct an Artinian quotient of the Stanley--Reisner ring of the link of the negative support. Its Hilbert series gives the graded dimensions of the orthogonal complement of the local fake indicial ideal and, under the Okuyama--Saito Frobenius condition, those of the leading logarithmic coefficient space of actual series solutions. The construction requires no Cohen--Macaulay hypothesis. When a top-dimensional standard pair occurs and the link is Cohen--Macaulay, the Hilbert series specializes to the $h$-polynomial of the link.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

10 extracted references · 9 canonical work pages

[6]

Okuyama and M

G. Okuyama and M. Saito,Logarithmic A-hypergeometric series II, Beitr. Algebra Geom.64(2023), no. 4, 1057–1086

work page 2023
[5]

Nagamine, A-hypergeometric Series with Parameters in the Core, Math

M. Nagamine, A-hypergeometric Series with Parameters in the Core, Math. J. Okayama Univ., to appear

work page
[1]

Bj¨ orner and M

A. Bj¨ orner and M. L. Wachs,Shellable nonpure complexes and posets. II, Trans. Amer. Math. Soc.349(1997), no. 10, 3945–3975

work page 1997
[2]

I. M. Gel’fand, M. M. Kapranov, and A. V. Zelevinsky,Discriminants, Resultants, and Multidimensional Determinants, Birkh¨ auser, Boston, 1994

work page 1994
[3]

Hotta,Equivariant D-modules, in: Proceedings of the ICPAM Spring School in Wuhan, 1991, available as arXiv:math/9805021

R. Hotta,Equivariant D-modules, in: Proceedings of the ICPAM Spring School in Wuhan, 1991, available as arXiv:math/9805021

arXiv 1991
[4]

Miller and B

E. Miller and B. Sturmfels,Combinatorial Commutative Algebra, Grad- uate Texts in Mathematics, vol. 227, Springer, New York, 2005

work page 2005
[7]

Saito,Logarithm-free A-hypergeometric series, Duke Math

M. Saito,Logarithm-free A-hypergeometric series, Duke Math. J.115 (2002), no. 1, 53–73

work page 2002
[8]

Saito,Logarithmic A-hypergeometric series, Int

M. Saito,Logarithmic A-hypergeometric series, Int. J. Math.31(2020), no. 13, 2050110

work page 2020

Show all 10 references

[9]

Saito, B

M. Saito, B. Sturmfels, and N. Takayama,Gr¨ obner Deformations of Hypergeometric Differential Equations, Algorithms and Computation in Mathematics, vol. 6, Springer, Berlin, 2000

2000
[10]

Schulze and U

M. Schulze and U. Walther,Irregularity of hypergeometric systems via slopes along coordinate subspaces, Duke Math. J.142(2008), no. 3, 465–509. Mao NAGAMINE Department of Mathematics Graduate School of Science Hokkaido University Sapporo 060-0810, JAPAN e-mail: nagamine.mao.d9...

2008

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [6]

Okuyama and M

G. Okuyama and M. Saito,Logarithmic A-hypergeometric series II, Beitr. Algebra Geom.64(2023), no. 4, 1057–1086

work page 2023

[2] [5]

Nagamine, A-hypergeometric Series with Parameters in the Core, Math

M. Nagamine, A-hypergeometric Series with Parameters in the Core, Math. J. Okayama Univ., to appear

work page

[3] [1]

Bj¨ orner and M

A. Bj¨ orner and M. L. Wachs,Shellable nonpure complexes and posets. II, Trans. Amer. Math. Soc.349(1997), no. 10, 3945–3975

work page 1997

[4] [2]

I. M. Gel’fand, M. M. Kapranov, and A. V. Zelevinsky,Discriminants, Resultants, and Multidimensional Determinants, Birkh¨ auser, Boston, 1994

work page 1994

[5] [3]

Hotta,Equivariant D-modules, in: Proceedings of the ICPAM Spring School in Wuhan, 1991, available as arXiv:math/9805021

R. Hotta,Equivariant D-modules, in: Proceedings of the ICPAM Spring School in Wuhan, 1991, available as arXiv:math/9805021

arXiv 1991

[6] [4]

Miller and B

E. Miller and B. Sturmfels,Combinatorial Commutative Algebra, Grad- uate Texts in Mathematics, vol. 227, Springer, New York, 2005

work page 2005

[7] [7]

Saito,Logarithm-free A-hypergeometric series, Duke Math

M. Saito,Logarithm-free A-hypergeometric series, Duke Math. J.115 (2002), no. 1, 53–73

work page 2002

[8] [8]

Saito,Logarithmic A-hypergeometric series, Int

M. Saito,Logarithmic A-hypergeometric series, Int. J. Math.31(2020), no. 13, 2050110

work page 2020