Pith. sign in

REVIEW 3 major objections 5 minor 3 references

Explicit constructions of short virtual resolutions of truncations

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read For every nef divisor on a smooth projective toric variety, the paper constructs an explicit cellular free resolution of a ceiling truncation ideal, of length at most the dimension, and shows it matches a symplectic-geometry construction.

desk verdict A valuable explicit construction of ceiling truncation resolutions on toric varieties, with a genuine gap in the isomorphism theorem that needs patching before the main comparison is airtight. read the letter →

arxiv 2501.06960 v2 pith:7NDMTAPT submitted 2025-01-12 math.AG math.AC

Lauren Cranton Heller This is my paper · ORCID
classification math.AGmath.AC MSC 13D0214M25
keywords toricvarietiescellularresolutionsceilingtruncationshortvirtualFourier-MukaitransformssyzygiestotalcoordinateringsEagon-Northcottcomplexes
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 gives every nef divisor on a smooth projective toric variety an explicit free resolution of a naturally defined monomial ideal, with length no more than the dimension of the variety. The ideal, called the ceiling truncation of the total coordinate ring, is generated by the ceiling labels of a finite cell complex built from the polytope of global sections. The complex is shown to be a cellular resolution, and after twisting by the divisor it reproduces the Fourier–Mukai transform of $\mathcal O(d)$ along a known resolution of the diagonal. The same combinatorial truncation extends to arbitrary modules and computes the homology of the transformed modules; the author exhibits a smooth toric variety where this homology is nonzero for every positive multiple of an ample divisor, so exactness can fail at the level of modules. If correct, the construction is the missing toric analogue of the Eagon–Northcott resolution of powers of the maximal ideal in projective space.

What carries the argument

The engine is the alcove cell structure obtained by subdividing $\mathbb R^r$ along all integral translates of the coordinate hyperplanes, then restricting to the affine plane $\alpha + M\otimes\mathbb R$ and to the positive quadrant, producing the finite labeled cell complex $D$. Each cell is labeled by rounding fractional exponents upward to the nearest integer, and a lemma shows that every cell label is the least common multiple of its boundary labels; the ideal generated by these labels is the ceiling truncation $\operatorname{trunc}_d(S)$. The complex $F_\bullet$ is the cellular complex of $D$ with respect to an incidence function. The isomorphism with the Fourier–Mukai transform rests on a bijection between open cells of $D$ lying over a torus cell and monomials of degree $d+\deg x^{\lfloor\gamma\rfloor}$, which unfolds the torus cell complex into the polytope of sections.

What would settle it

Choose a smooth toric variety where the quotient cell complex $E$ identifies multiple preimages of a cell (as in the worked example), fix two incidence functions on $C$ that are invariant under translation by $M$, compute the boundary maps of $F(d)_\bullet$ for both, and compare their homologies; if the homologies differ, or if $\partial^2\neq 0$ for either one, the central resolution and isomorphism claims fail.

Watch

Extended reading notes

Core claim

The central claim is that for every nef divisor $d$ on a smooth projective toric variety $X$, the monomial ideal $\operatorname{trunc}_d(S)$ generated by the ceiling labels of an explicitly defined labeled cell complex $D$ inside the polytope of sections of $\mathcal O(d)$ has a cellular free resolution $F_\bullet$ of length at most $\dim X$. After twisting by $d$, this same complex sheafifies to the Fourier–Mukai transform of $\mathcal O(d)$ with respect to a recently constructed resolution of the diagonal, giving an explicit algebraic model of that transform for structure sheaves of nef divisors. The construction also yields a truncation functor on modules: for a module $Q$ satisfying a nefness condition, the homology of the ceiling-truncated free resolution of $Q$ computes the homology of the Fourier–Mukai transform of $Q(d)$. A notable consequence is that this homology need not vanish: on an explicit smooth toric variety, the transform of $Q(md)$ has nontrivial homology for all sufficiently large multiples $m$, so exactness can fail at the module level even when the sheafified complex is exact.

Load-bearing premise

The load-bearing premise is that a choice of incidence function can be made globally on the infinite cell complex, invariant under the lattice, so that the boundary maps on the finite cell complex $D$ are exactly the same as the maps in the Fourier–Mukai complex on the quotient torus; the paper checks this locally but not globally, and cells in the quotient can have multiple preimages with identified boundaries.

Editorial extensions

If this is right

  • For each nef divisor $d$, $\operatorname{trunc}_d(S)$ is resolved in length at most $\dim X$ by an explicitly described cellular complex, giving a concrete algebraic model for the Fourier–Mukai transform of $\mathcal O(d)$.
  • The sheafification of the resolution is exactly $\mathcal O(d)$, so the complex is a short virtual resolution in the sense that it has no sheaf-level homology beyond the structure sheaf.
  • The homology of the Fourier–Mukai transform $\Phi(Q(d))$ for any suitable module $Q$ can be read off from the ceiling-truncated free resolution of $Q$, as stated in Corollary 6.5.
  • The nontrivial homology exhibited in Example 6.7 shows that exactness can fail at the module level while the sheafified complex is exact, supporting the broader philosophy that allowing irrelevant homology enlarges the class of usable resolutions.

Reading between the lines

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

  • The construction suggests a practical algorithmic route to short virtual resolutions of line bundles on any smooth projective toric variety, avoiding spectral sequences and mirror functors, and it seems likely to be implementable directly from the fan data.
  • The persistent homology in Example 6.7 may be a combinatorial shadow of derived autoequivalences or of multigraded Castelnuovo–Mumford regularity, and could be studied purely through the alcove subdivision of the section polytope.
  • One could test whether the same ceiling-truncation construction extends to arbitrary coherent sheaves by applying it to a chosen free resolution; Corollary 6.5 already predicts the resulting homology, so the open question is whether a canonical choice of resolution makes the truncation functor exact in more cases.
  • The dependence on the incidence function suggests that non-unimodular fans may carry an orientation-type obstruction; comparing two invariant incidence functions on a Picard-rank-two example would isolate whether the main isomorphism is independent of that choice.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces a notion of ceiling truncation for the total coordinate ring S of a smooth projective toric variety X, and constructs, for each d in the nef cone, a finite cellular complex F_• supported on the polytope of sections of O(d). Theorem 2.5 asserts that F_• is a free resolution of the ceiling truncation truncd(S), with length at most dim X. In Section 5 the author claims that after twisting by d, F(d)_• is isomorphic to the Fourier-Mukai transform of O(d) with respect to the Hanlon-Hicks-Lazarev resolution of the diagonal. Section 6 defines a truncation functor for arbitrary modules and uses it, via a spectral sequence argument, to compare homology of the transform with homology of the truncated resolution; Example 6.7 reports nontrivial homology for all positive multiples of an ample class. The paper also contains a fully worked example for the Hirzebruch surface F2.

Significance. If the main comparison theorem is fully established, the paper gives a valuable explicit and computable analogue of Eagon-Northcott complexes for nef truncations on all smooth projective toric varieties, and it connects commutative algebra with the symplectically motivated short virtual resolutions of Hanlon, Hicks, and Lazarev. The convexity proof of Theorem 2.5 is clean and appears correct, and the use of Macaulay2 in Section 4 is a useful computational check. The claimed nontrivial homology in Section 6 would be a genuinely new phenomenon if supported by a complete argument. However, the map-level comparison in Theorem 5.1 and the degeneration claim in Corollary 6.5 are not fully proved as written, and Example 6.7 does not contain the promised hand-checkable proof for all multiples. These issues are load-bearing for the paper's central claims and require revision.

major comments (3)
  1. [§5, proof of Theorem 5.1; §3, (3.1)] The proof of Theorem 5.1 asserts an isomorphism of complexes F(d)_• ≅ Φ(S(d)) using 'incidence functions from the same ε on C', but no such globally M-invariant incidence function is constructed. Lemma 5.2 establishes only a one-to-one correspondence between open cells of D and monomials; it does not compare boundary maps when cells of E have multiple preimages in C. The text immediately before (3.1) explicitly says that no global set of representatives can be chosen so that every boundary incidence in E lifts to a boundary incidence in C, and Section 4 displays a horizontal edge in E whose two ends are identified to the same vertex. For a map-level isomorphism one must prove that a single ε on C is invariant under translation by M and induces the same cellular boundary maps on D and on E, including all preimages and sign choices. As written, only one incidence pair is checked, so the isomorphism, and with it Corollaries 5.3 and 6.5, is not fully established. Please either construct ε explicitly, prove its existence by a global orientation argument, or revise the statement to a weaker comparison that does not require a global incidence function.
  2. [§6, Example 6.7] Example 6.7 states that 'homology can be exhibited by hand for all m' and identifies generators arising from the term R(1,-1,1,-1) in G1, but no calculation is given. This is load-bearing: the claimed persistence of homology for all multiples of an ample divisor is the paper's main new phenomenon, stated in the final paragraph of Section 6. A Macaulay2 check for finitely many m does not establish the infinite family. Please include the hand calculation or provide a complete supplementary argument proving the stated vanishing failure for every m.
  3. [§6, proof of Corollary 6.5] The proof of Corollary 6.5 contains the sentence 'In the spectral sequence of the double complex all maps are zero starting at the second page for both directions' with no justification. The degeneration of this spectral sequence is exactly what identifies the homology of truncd(K_•)(d) with that of Φ(Q(d)); without a proof or a reference to a proved lemma, the corollary is not established. Please supply the computation of the second page or state the missing hypothesis that makes the degeneration immediate.
minor comments (5)
  1. [Introduction, §4] There is a typographical error: 'analagous' should be 'analogous' in the sentence about pictures from symplectic geometry.
  2. [Figure 5 and surrounding text] Figure 5 is difficult to parse: the caption lists 'edges in D F1 G1 Φ(S(d))1 polytope in S' without indicating which columns correspond to which complex. Please restructure the figure or caption so the correspondence of summands is explicit.
  3. [Remark 6.4] Remark 6.4 says 'the author believes that a general definition is possible'; this speculative statement should be removed or replaced by a precise conjecture, since it is not used in the paper.
  4. [Section 4] The incidence function on D is said to be 'chosen by Macaulay2'; for reproducibility, please provide the code or explicitly list the orientations of all cells in Figures 3 and 4.
  5. [References] Reference [And23] is listed as 'Pre-published' with a 2024 arXiv number; please update its publication or arXiv status.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper's construction is self-contained and its main comparison is against an independent external construction.

full rationale

The derivation chain is not circular. Construction 2.2 and Theorem 2.5 build a cellular resolution of truncd(S) directly from the polytope D, using only the Bayer–Sturmfels criterion; no fitted parameter or target conclusion is fed into the construction. Theorem 5.1 compares F(d)• with the Fourier–Mukai transform of S(d) using the resolution of the diagonal of Hanlon, Hicks, and Lazarev, which is an independent construction cited from [HHL24]; the isomorphism is argued by a term-by-term bijection in Lemma 5.2 and a matching of differentials using the same incidence function. Even though the global existence of an M-invariant incidence function on C is not fully proved and is a genuine gap, that is a correctness or completeness issue, not circularity. The only self-citation is [BCHS22], which is contextual background on multigraded regularity and is not load-bearing for the main theorems. Consequently, the central claims do not reduce to their inputs by definition or by self-citation, and the appropriate circularity score is 0.

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

The central arguments depend on standard results in cellular resolutions and toric geometry, on the Hanlon-Hicks-Lazarev diagonal resolution, and on vanishing theorems for nef line bundles. No numerical parameters are fitted; the only inputs are the toric fan and the chosen nef divisor d.

assumptions (6)
  • standard math Bayer-Sturmfels criterion: a labeled cell complex is a cellular resolution of a monomial ideal if each induced subcomplex on labels dividing x^beta is contractible.
    Invoked in the proof of Theorem 2.5 via [BS98, Proposition 1.2].
  • standard math Vanishing theorem for cohomology of nef line bundles on toric varieties: H^i(X, O(deg x^{floor(alpha+gamma)})) = 0 for i > 0.
    Used in Lemma 3.2 to force the Fourier-Mukai spectral sequence to collapse to a single row; cites [CLS11, Theorem 9.3.5].
  • standard math The Hanlon-Hicks-Lazarev resolution of the diagonal is a resolution of the structure sheaf of the diagonal on X times X, with terms described in Section 3.
    External theorem [HHL24] on which Theorem 5.1 and the Fourier-Mukai comparison rest.
  • standard math Cox's correspondence between finitely generated graded S-modules and coherent sheaves on X, including B-saturation behavior.
    Used in Corollary 5.3 to relate sheafification of the resolution to O(d); cites [Cox95, Theorem 3.7].
  • domain assumption The smooth projective toric variety X is defined over an algebraically closed field by a full-dimensional fan with primitive ray vectors, and Pic X is torsion-free.
    Set at the start of Section 1; the construction uses the quotient M tensor R over M and the real cell structure C.
  • ad hoc to paper The spectral sequence of the double complex in Corollary 6.5 has all maps zero starting at the second page in both directions.
    Stated in the proof of Corollary 6.5 without full detail; it is load-bearing for the homology comparison between the ceiling truncation and the Fourier-Mukai transform.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Explicit constructions of short virtual resolutions of truncations." pith.science (2026). https://pith.science/paper/7NDMTAPT

@misc{pith2026250106960,
  author       = {Pith},
  title        = {Pith review of: Explicit constructions of short virtual resolutions of truncations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7NDMTAPT}},
  note         = {Machine review of arXiv:2501.06960}
}
read the original abstract

We propose a concept of truncation for arbitrary smooth projective toric varieties and construct explicit cellular resolutions for nef truncations of their total coordinate rings. We show that these resolutions agree with the short resolutions of Hanlon, Hicks, and Lazarev, which were motivated by symplectic geometry, and we use our definition to exhibit nontrivial homology in the commutative algebraic analogue of their construction.

Figures

Figures reproduced from arXiv: 2501.06960 by the authors.

Figure 1
Figure 1. the fan of X and the degrees of the variables of S They form the presentation for Pic X ≃ Z 2 as follows: 0 /M      1 0 0 1 −1 2 0 −1      /Z 4 " 1 −2 1 0 0 1 0 1# /Pic X /0 (4.1) Fix d = (1, 1) and α = (0, 0, 1, 1), corresponding to the monomial zw. All monomials of degree d fit into the trapezoid in [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. the polytope of degree d monomials in S The restrictions of the coordinate planes in R 4 are then given by the hyperplanes where ⟨−, uρ⟩ is integer valued. The labeled cell complex D is pictured in [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. the labeled CW complex D Remark 4.1. In this example Sd generates all of truncd(S), which is not true in general, yet unfortunately common to examples of a reasonable size for hand computation. We construct the complex F• from D (and an incidence function chosen by Macaulay2 ). By Lemma 2.3 the remaining labels are the least common multiples of those on the vertices. Refer to [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: the CW complex E and its labeled covering space ∂1 = h x ⊗ z − z ⊗ x xy ⊗ w yz ⊗ w w ⊗ xy w ⊗ yz 0 −1 ⊗ x −1 ⊗ z −x ⊗ 1 −z ⊗ 1 i , ∂2 =   −y ⊗ w 0 −w ⊗ y 1 ⊗ z −z ⊗ 1 0 −1 ⊗ x x ⊗ 1 0 0 −1 ⊗ z −z ⊗ 1 0 1 ⊗ x x ⊗ 1   For instance, in the first entry of the first mat…
Figure 5
Figure 5. Figure 5: the correspondence of summands between F1 and G1 Proof. We construct a bijection. Given an open cell σ ⊂ D ⊂ R r whose image contains γ +M let β ∈ σ be a preimage of γ + M and consider the monomial x ⌊β⌋ . It is well defined because β ≥ 0 by definition of D and ⌊β⌋ ̸= …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

3 extracted references · 2 canonical work pages

  1. [111]

    14 REFERENCES Department of Mathematics, University of Nebraska, 203 A very Hall, Lincoln, NE 68588 Email address : lheller2@unl.edu URL: https://lcrantonh.github.io/

    doi: 10.1090/bproc/72. 14 REFERENCES Department of Mathematics, University of Nebraska, 203 A very Hall, Lincoln, NE 68588 Email address : lheller2@unl.edu URL: https://lcrantonh.github.io/

    work page doi:10.1090/bproc/72

  2. [760]

    Resolutions of toric subvarieties by line bundles and applications

    doi: 10.1016/j.geomphys.2008.01.004. [CLS11] David A. Cox, John B. Little, and Henry K. Schenck. Toric Varieties. Vol. 124. Graduate Studies in Mathematics. American Mathematical Society, Providence, RI, 2011. xxiv+841. doi: 10.1090/gsm/124. [Cox95] David A. Cox. “The Homogeneous Coordinate Ring of a Toric Variety”. In: Journal of Algebraic Geometry 4.1 (...

    work page Pith review arXiv doi:10.48550/arxiv.2303.03763 1995

  3. [2006]

    Multigraded Castelnuovo-Mumford Regularity

    viii+307. doi: 10.1093/acprof:oso/9780199296866.001.0001. [M2] Daniel R. Grayson and Michael E. Stillman. Macaulay2, a Software System for Research in Algebraic Geometry. url: http://www2.macaulay2.com. [MS04] Diane Maclagan and Gregory G. Smith. “Multigraded Castelnuovo-Mumford Regularity”. In: Journal f¨ ur die Reine und Angewandte Mathematik. [Crelle’s...

    arXiv 2004

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.