Pith. sign in

REVIEW 2 major objections 5 minor 33 references

A torsion theoretic interpretation for sheaves of modules and Grothendieck topologies on directed categories

T0 review · 2 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Sheaves of modules are exactly the torsion-saturated presheaves.

desk verdict A solid, important torsion-theoretic characterization of sheaves and a noetherian rigidity theorem, marred by a few genuine but fixable errors (abstract typo, wrong shape claim in Prop 6.3, and a misapplied theorem in Example 7.14). read the letter →

arxiv 2506.08685 v3 pith:M76ZHZG2 submitted 2025-06-10 math.RT math.CT

Zhenxing Di , Liping Li , Li Liang This is my paper · ORCID
classification math.RTmath.CT MSC 18F1018E4018E3518G10
keywords GrothendiecktopologiessheavesofmodulestorsionpairsSerrequotientsEIcategoriesrigidrepresentationstabilitydirected
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

The paper proves that sheaf theory on a ringed site can be recast as torsion theory in the category of presheaves of modules. For any Grothendieck topology J, a presheaf of modules is a sheaf exactly when it is J-saturated: it has no J-torsion submodule and its first derived torsion functor vanishes. Equivalently, a sheaf is a module right perpendicular to every J-torsion module, meaning both Hom and $Ext^{1}$ to torsion modules vanish. The sheaf category is therefore the Serre quotient of the category of all O-modules by the J-torsion modules. The paper also shows that on directed categories with finite endomorphism monoids, all Grothendieck topologies are rigid whenever the category is noetherian and EI, which makes every sheaf category equivalent to a presheaf category over a full subcategory; it then classifies all topologies on a large family of such categories and transfers finiteness and injectivity results from FI and VI to their infinite full subcategories.

What carries the argument

The load-bearing object is the torsion functor T_J, which sends an O-module V to the submodule generated by all elements killed by some covering sieve in J; an element v in V_x is J-torsion when a covering sieve S ∈ J(x) sends v to 0. The saturation condition T_J(V)=0 and $R^{1}$T_J(V)=0 is what characterizes sheaves, and it is shown to be equivalent to being right perpendicular to all J-torsion modules. For the rigidity results, the carrying mechanism is the family of minimal covering sieves S_x: a Grothendieck topology on a noetherian EI category is determined by a consistent family satisfying S_x = ⋃_{y≠x} S_y ∘ C(x,y), and rigidity means each S_x is generated by morphisms to J-irreducible objects.

What would settle it

On a one-object directed category whose endomorphism monoid is the monoid of all surjections of an infinite set, check directly whether the dense topology and the maximal topology are the only two Grothendieck topologies; if a third topology exists, the finite-monoid dichotomy in Lemma 5.9 breaks down and the necessity direction of the rigidity theorem would fail without the finiteness condition.

Watch

Extended reading notes

Core claim

The central claim is a homological characterization of sheaves of modules on a ringed site: an O-module V is a sheaf if and only if it is J-saturated, meaning T_J(V)=0 and $R^{1}$T_J(V)=0, where T_J is the left exact endofunctor sending V to its maximal J-torsion submodule. This is equivalent to V being right perpendicular to every J-torsion module, and it yields the equivalence Sh(C^op,O) ≃ O-Mod / T(J). The paper further claims that for a directed category C whose endomorphism monoids are all finite, every Grothendieck topology on C^op is rigid if and only if C is a noetherian EI category, so sheaf categories reduce to presheaf categories over the full subcategory of J-irreducible objects. For EI categories of type N or Z, all Grothendieck topologies are explicitly classified by sequences d satisfying d(n) ≠ 0 implies d(n+1) = d(n) - 1, and non-rigid topologies are almost atomic.

Load-bearing premise

The necessity direction of the rigidity classification assumes every object has finitely many endomorphisms, because the proof needs the fact that a finite monoid with exactly two Grothendieck topologies is a group; without that finiteness the characterization can fail, though the sheaf characterization itself does not depend on it.

Editorial extensions

If this is right

  • Injective objects in the sheaf category are exactly the J-torsion-free injective O-modules.
  • Sheaf cohomology groups can be computed as R^iΓ_x ≅ Γ^p_x ∘ R^{i+1}T_J, so cohomology of sheaves is controlled by the derived functors of the torsion functor.
  • Sheafification has an elementary two-step description: take the torsion-free quotient, embed it into an injective hull, and pull back the torsion part of the cokernel.
  • For noetherian EI categories, every sheaf category is equivalent to a presheaf category over the full subcategory of J-irreducible objects, and J-torsion modules are precisely modules supported outside that subcategory.
  • For EI categories of type N or Z, all Grothendieck topologies are classified by stepping sequences, and every non-rigid topology is almost atomic, reducing sheaf questions to atomic-site questions.
  • Every finitely generated module over an infinite full subcategory of FI or VI_q is noetherian, and over characteristic-zero fields the projective-injective, finite-injective-dimension, and Serre-quotient properties hold.

Reading between the lines

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

  • The Serre-quotient description gives a practical recipe the paper leaves implicit: to study sheaves on a ringed site, one can work entirely inside O-Mod and compute R^1T_J directly rather than constructing injective resolutions in the sheaf category.
  • The classification by sequences d suggests a testable invariant: two topologies on a type-N or type-Z category might give equivalent sheaf categories exactly when their d-functions agree on a cofinite tail; the paper does not state this, but it is consistent with the almost-atomic description.
  • One could use the torsion-theoretic characterization to decide whether the irreducible sheaves constructed from weights in group representations are inequivalent; if they are, the paper's closing questions give a direct route toward a bijective formulation of Alperin's weight conjecture.
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

2 major / 5 minor

Summary. The paper introduces a torsion-theoretic framework for sheaves of modules on a ringed site (C^op, O). For a Grothendieck topology J, the authors define a J-torsion subfunctor T_J on O-Mod and prove that (T(J), F(J)) is a hereditary torsion pair; they then characterize sheaves of modules as J-saturated O-modules, equivalently as modules right perpendicular to all J-torsion modules (Theorem 1.1). This yields the Serre-quotient equivalence Sh(C^op, O) ≃ O-Mod/T(J) (Corollary 1.2). In the second part, the paper studies Grothendieck topologies on directed categories: Theorem 1.4 asserts that, when all endomorphism monoids are finite, every topology on C^op is rigid if and only if C is a noetherian EI category. The paper also classifies all topologies on EI categories of type N and Z (Theorem 1.7) and derives applications to infinite full subcategories of FI and VI_q, including noetherianity, local self-injectivity, and Serre-quotient descriptions (Theorem 1.8).

Significance. The homological characterization in Theorem 1.1 is a substantial and useful bridge between sheaf theory and representation theory; it extends the authors' earlier atomic-site result [8] to arbitrary Grothendieck topologies and gives an explicit description of the localizing subcategory in the sheafification localization. The proofs of the central theorem are detailed and internally consistent, and the paper is careful to state the finiteness assumption in Theorem 1.4 and to note in Remark 5.14 where it is used. The classifications and applications to FI/VI_q are interesting and give falsifiable, checkable statements. The main reservation is that two non-central claims—the 'more explicitly' decomposition in Proposition 6.3 and the application in Example 7.14—are incorrect as stated and need correction.

major comments (2)
  1. [§7.2, Example 7.14] The example asserts that for the orbit category of an artinian group such as the Prüfer p-group, 'by Theorem 5.7 every Grothendieck topology J on it is rigid.' Theorem 5.7, however, gives this conclusion only for noetherian EI categories (under the finite-endomorphism hypothesis); its necessary direction even shows that if every topology is rigid then C is noetherian. An artinian EI category that is not noetherian (the Prüfer p-group orbit category is such) need not have all topologies rigid, and the dense topology is a natural counterexample. This application should be re-examined and either restricted to a noetherian setting or proved directly.
  2. [§6, Proposition 6.3] The 'More explicitly' description of the sequences d is false. The function d(2k)=1, d(2k+1)=0 satisfies the condition 'if d(n)≠0 then d(n+1)=d(n)−1' and therefore defines a generic Grothendieck topology, but the sequence 1,0,1,0,... cannot be written as a block of zeros followed by at most one finite descending block (r,r−1,...,1,0) and then an optional ∞ tail. The parameterization by functions d is correct, but the claimed normal form and its analogue in Proposition 6.8 must be corrected or removed.
minor comments (5)
  1. [Throughout] There are several typographical errors, including 'Noth that Sy = C(y, −)' in the proof of Corollary 5.17, 'T ransitivity' in the proof of Proposition 6.3, 'wight' for 'weight' near the end of the paper, and 'Pr¨ uferp-group' missing a space in Example 7.14.
  2. [Abstract and §1] The abstract and introduction describe Theorem 1.8 as extending properties of 'F and VI', but the theorem itself states FI and VI_q; please make the notation consistent.
  3. [§6] The term 'generic Grothendieck topology' is used informally in Section 6; a formal definition would improve clarity, since the term appears in Proposition 6.3 and Corollary 6.7.
  4. [§6, Proposition 6.6] The proof of Proposition 6.6 is omitted with the note that it is similar to Proposition 6.3; for the non-generic case, the transitivity axiom requires a case analysis involving the empty sieve, so at least a sketch of that check would be helpful.
  5. [§7.1] The dense topology is denoted Jd in Section 2 and also appears as Jd in Section 7.1 after Lemma 7.1; these two uses could be confused, and a different symbol for one of them would be preferable.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity; the central sheaf/torsion equivalence is proved from the Grothendieck-topology axioms, with self-citations to the authors' prior atomic-site paper [8] only supplying special-case strategy and auxiliary results.

full rationale

The core derivation chain is self-contained. Section 3 builds the torsion functor T_J directly from the covering sieves and proves, from the stability and transitivity axioms, that (T(J), F(J)) is a hereditary torsion pair; Proposition 3.8's bijection between topologies and torsion classes is proved via the annihilator rule A(T(J)), not assumed. Theorem 4.2 proves sheaf iff J-saturated iff right perpendicular using the injective-hull and sheafification arguments in the text; the citation to [8, Lemma 3.6] is only a strategy pointer for the atomic special case, and the general injective-hull argument is written out. Corollary 1.2's Serre-quotient equivalence is imported from the external Geigle-Lenzing theory [15], not from the authors' own work. Section 5's rigidity theorem is proved with Lemma 5.9 and the consistent-family construction, and Section 6's classification is derived from the axioms via the functions d; the false explicit decomposition in Proposition 6.3 and the questionable application of Theorem 5.7 in Example 7.14 are correctness concerns, not circularity, because the parameterization itself is not assumed from the conclusion. The remaining self-citations to [8] (Corollaries 4.4, 4.6, sheafification details, Lemma 7.1, and parts of the Theorem 1.8 proof) cite prior work for special cases or for known results about the ambient categories C; none of them assumes the target theorem for infinite full subcategories D. The paper also transparently states the finiteness limitation in Remark 5.14. Accordingly, no derivation step reduces to its own input; the only reason the score is not 0 is the recurring reliance on the authors' own [8] for supporting lemmas, which is not load-bearing circularity.

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

The paper's claims rest on standard set theory, Grothendieck category theory, and specific structural assumptions about the categories studied. No free parameters or invented entities appear; the novel concepts (J-torsion, J-saturated, consistent families) are defined directly from the input topology, not postulated independently.

assumptions (6)
  • standard math The ambient metatheory is ZFC with standard set-theoretic foundations, including the existence of Grothendieck universes when referring to small and large categories.
    Required for topos-theoretic statements such as categories of sheaves and Grothendieck categories.
  • domain assumption For a presheaf of rings O on a small category C, the category O-Mod is a Grothendieck category with enough injectives.
    Used in the proof of Theorem 4.2 to take injective hulls and derived functors; standard for ringed topoi.
  • domain assumption In Theorem 1.4, C is a directed category with finite endomorphism sets C(x,x).
    The finite endomorphism condition is essential for Lemma 5.9 and is acknowledged in Remark 5.14 as not removable for the only-if direction.
  • domain assumption In Theorem 1.7, C is an EI category of type N or Z, defined by objects N or Z, C(m,n) nonempty iff m is at most n, a transitive action, a rank function, and generation in degrees 0 and 1.
    These combinatorial restrictions make the classification of Grothendieck topologies tractable; the type Z proof is only sketched.
  • domain assumption In Section 7, k is a commutative noetherian ring, and for the injectivity results a field of characteristic 0; the categories FI and VI_q are assumed.
    The applications require established noetherianity and injectivity results for these categories cited from [7,12,25,27,28].
  • standard math The Grothendieck topologies under consideration satisfy the three axioms (maximal, stability, transitivity) as defined in Definition 2.1.
    This is the standard definition from Mac Lane-Moerdijk, taken as given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A torsion theoretic interpretation for sheaves of modules and Grothendieck topologies on directed categories." pith.science (2026). https://pith.science/paper/M76ZHZG2

@misc{pith2026250608685,
  author       = {Pith},
  title        = {Pith review of: A torsion theoretic interpretation for sheaves of modules and Grothendieck topologies on directed categories},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M76ZHZG2}},
  note         = {Machine review of arXiv:2506.08685}
}
abstract

We prove that every Grothendieck topology induces a hereditary torsion pair in the category of presheaves of modules on a ringed site, and obtain a homological characterization of sheaves of modules: a presheaf of modules is a sheaf of modules if and only if it is saturated with respect to torsion presheaves, or equivalently, it is right perpendicular to torsion presheaves in the sense of Geigle and Lenzing. We also study Grothendieck topologies on directed categories $\mathscr{C}$ satisfying certain finiteness condition, and show that every Grothendieck topology on $\mathscr{C}$ is a subcategory topology if and only if $\mathscr{C}$ is an artinian EI category. Consequently, in this case every sheaf category is equivalent to the presheaf category over a full subcategory of $\mathscr{C}$. Finally, we classify all Grothendieck topologies on a special type of noetherian EI categories, and extend the locally self-injective property of representations of $\mathrm{F}$ and $\mathrm{VI}$ to representations of their infinite full subcategories. Some potential applications in group representation theory are given at the end of this paper.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 29 canonical work pages

  1. [8]

    Z. Di, L. Li, L. Liang, and F. Xu. Sheaves of modules over atomic sites and discrete representations of topological groups. To appear in Trans. Amer. Math. Soc. arXiv:2108.13600

    work page Pith review arXiv

  2. [1]

    M. Artin. Grothendieck topologies. Lecture Notes, Harvard Univ., 1962

    work page 1962

  3. [2]

    Artin, A

    M. Artin, A. Grothendieck, and J. Verdier. Th´ eorie de topos et cohomologie ´ elale des sch´ emas. Lect. Notes Math. 269 and 270, Springer-Verlag, 1972

    work page 1972

  4. [3]

    P. Balmer. Modular representations of finite groups with trivial restriction to Sylow subgroups, J. Eur. Math. Soc. 15 (2013), 2061-2079

    work page 2013

  5. [4]

    S. Bazzoni. Ring Epimorphisms, Gabriel Topologies and Contramodules. In: M. Clementino, A. Facchini, M. Gran (eds) New Perspectives in Algebra, Topology and Categories. Coimbra Mathematical Texts, vol 1. Springer, Cham

    work page

  6. [5]

    Enveloping Classes over Commutative Rings

    S. Bazzoni and G. Le Gros. Envolping classes over commutative rings. Preprint. ArXiv:1901.07921

    work page Pith review arXiv 1901

  7. [6]

    Churhch, J

    T. Churhch, J. Ellenberg, and B. Farb. FI-modules and stability for representations of symmetric groups. Duke Math. J. 164 (2015), 1833-1910

    work page 2015

  8. [7]

    Church, J

    T. Church, J. Ellenberg, B. Farb, and R. Nagpal. FI-modules over noetherian rings. Geom. Topol. 18 (2014), 2951-2984. arXiv:1210.1854. A TORSION THEORETIC INTERPRETATION FOR SHEA VES OF MODULES 35

    arXiv 2014

Show all 33 references
  1. [9]

    Z. Di, L. Li, L. Liang, and N. Yu. Representations over diagrams of abelian categories I: Global structure and homological objects. J. Algebra 672 (2025), 208-246. arXiv:2210.08558

    2025 arXiv

  2. [10]

    P. Gabriel. Des cat´ egories ab´ eliennes. Bull. Soc. Math. France 90 (1962), 323-348

    1962

  3. [11]

    Gabriel and M

    P. Gabriel and M. Zisman. Calculus of fractions and homotopy theory. Springer-Verlag, 1967

    1967

  4. [12]

    W. L. Gan and L. Li. Coinduction functor in representation stability theory. J. London Math. Soc. 92 (2015) 689–711

    2015

  5. [13]

    W. L. Gan and L. Li. Adjoint functors on the representation category of OI-modules. Algebr. Represent. Theor. 26 (2023), 2015–2038

    2023

  6. [14]

    W. L. Gan, L. Li, and C. Xi. An application of Nakayama functor in representation stability theory. Indiana Univ. Math. J. 69 (2020), 2325-2338

    2020

  7. [15]

    Geigle and H

    W. Geigle and H. Lenzing. Perpendicular categories with applications to representations and sheaves. J. Algebra 144 (1991), 273-343

    1991

  8. [16]

    G¨ unt¨ urk¨ un and A

    S. G¨ unt¨ urk¨ un and A. Snowden. The representation theory of the increasing monoid. Preprint, arXiv:1812.12042v1

  9. [17]

    Johnstone

    P. Johnstone. Sketches of an elephant: a topos theory compendium, Vol. 1 and 2. The Clarendon Press, Oxford University Press, 2002

    2002

  10. [18]

    Hemelaer

    J. Hemelaer. Grothendieck topologies on posets. Preprint, arXiv:1811.10039

    arXiv

  11. [19]

    Kashiwara and P

    M. Kashiwara and P. Schapira. Categories and sheaves, Grundlehren der Mathematischen Wissenschaften 332. Springer-Verlag, 2006

    2006

  12. [20]

    A. Kock, G. Wraith. Elementary Toposes. Lecture Notes Series 30, Matematisk Institut, Aarhus Universitet, Aarhus, 1971

    1971

  13. [21]

    Li and E

    L. Li and E. Ramos. Depth and the local cohomology of FI G-modules. Adv. Math. 329 (2018), 704-741

    2018

  14. [22]

    Lindenhovius

    B. Lindenhovius. Grothendieck topologies on a poset. Preprint, arXiv:1405.4408

    arXiv

  15. [23]

    Mac Lane and I

    S. Mac Lane and I. Moerdijk. Sheaves in geometry and logic: a first introduction to topos theory, corrected reprint of the 1992 edition. Springer-Verlag, 1994

    1992

  16. [24]

    D. Murfet. Grothendieck topologies on quivers. Available at therisingsa.org

  17. [25]

    R. Nagpal. VI-modules in non-describing characteristic, part I. Algebra Number Theory 13 (2019), 2151-2189. arXiv:1709.07591

    2019 arXiv

  18. [26]

    R. Nagpal. VI-modules in non-describing characteristic, part II. J. reine angew. Math., 781 (2021), 187-205. arXiv:1810.04592

    2021 arXiv

  19. [27]

    S. Sam, A. Snowden, GL-equivariant modules over polynomial rings in infinitely many variables, Trans. Amer. Math. Soc. 368 (2016), 1097-1158. arXiv:1206.2233

    2016 arXiv

  20. [28]

    Sam and A

    S. Sam and A. Snowden. Gr¨ obner methods for representations of combinatorial categories. J. Amer. Math. Soc. 30 (2017), 159–203. arXiv:1409.1670

    2017 arXiv

  21. [29]

    Available at stacks.math.columbia.edu

    The Stacks Project. Available at stacks.math.columbia.edu

  22. [30]

    P. Webb. Standard stratifications of EI categories and Alperin’s weight conjecture. J. Algebra 320 (2008), 4073- 4091

    2008

  23. [31]

    Wu and F

    M. Wu and F. Xu. Skew category algebras and modules on ringed finite sites. J. Algebra 631 (2023), 194–217. arxiv:2207.04731

    2023 arXiv

  24. [32]

    Xiong and F

    T. Xiong and F. Xu. On sheaves in finite group representations. J. Pure Appl. Algebra 226 (2022), 107085. arXiv:2010.10372

    2022 arXiv

  25. [33]

    Xu and C

    F. Xu and C. Zheng. On cohomological characterizations of endotrivial modules. Preprint, arXiv:2308.16838. School of Mathematical Sciences, Huaqiao University, Quanzhou 362021, China Email address: dizhenxing@163.com LCSM(Ministry of Education), School of Mathematics and Stati...

    arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.