The derived $\infty$-category of Frobenius modules

Klaus Mattis; Timo Wei{\ss}

REVIEW 3 major objections 4 minor 26 references

The derived $\infty$-category of Frobenius modules

T0 review · 3 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read On any quasi-compact F_p-scheme with affine diagonal, the derived ∞-category of Frobenius modules is t-exactly equivalent to the category of Frobenius modules in the derived ∞-category of quasi-coherent sheaves.

desk verdict The main theorem is very likely correct and genuinely extends the authors' earlier work from regular Noetherian schemes to all geometric F_p-schemes, but the written proof leaves a load-bearing exactness premise for F_* unstated and even contradicted in the introduction. read the letter →

arxiv 2510.23267 v2 pith:K6V5A6SJ submitted 2025-10-27 math.AG math.AC

Klaus Mattis , Timo Wei{ss} This is my paper · ORCID

classification math.AGmath.AC MSC 14F3018G8018N60

keywords FrobeniusmodulesCartierderived∞-categoriesZariskidescentpositivecharacteristicquasi-coherentsheavest-structuresleft-complete

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 for a broad class of characteristic-p schemes — those that are quasi-compact and have affine diagonal, such as quasi-compact separated schemes — two natural ways of combining 'take the derived category' and 'impose a Frobenius action' produce the same stable ∞-category. The first construction derives the abelian category of Frobenius modules; the second takes Frobenius modules inside the derived category of quasi-coherent sheaves. The canonical comparison functor between them is shown to be a t-exact equivalence, so the homological t-structures agree as well. This removes the regularity and Noetherian hypotheses of an earlier result, and, as a byproduct, the paper establishes Zariski descent for the derived ∞-categories of Frobenius and Cartier modules. The strategy reduces the statement to affine schemes via descent and then identifies both sides with module spectra over a twisted polynomial ring.

What carries the argument

The central mechanism is the ∞-category Frob(C, G) of generalized Frobenius modules, defined as the lax equalizer of the identity functor and an endofunctor G; with C = QCoh(X) and G = F_*, the Frobenius pushforward, it recovers classical Frobenius modules. Its defining construction preserves limits, which makes Zariski descent possible. On affines, the proof relies on the module-category recognition theorem (an ∞-categorical version of Gabriel's theorem) to identify the two categories as module spectra over R[F]^op, and on a weak form of Grothendieck's AB4* axiom to obtain left-completeness of the quasi-coherent derived categories involved.

What would settle it

Take an affine singular F_p-algebra R such as R = F_p[x,y]/(xy) and compute the endomorphism ring spectrum of the rank-one Frobenius module (R, id: R → F_*R) inside D(Frob(Mod_R, F_*)); the theorem predicts this ring is the twisted polynomial ring R[F]^op, so any deviation would falsify the affine case and hence the main theorem.

Watch

Extended reading notes

Core claim

The central discovery is that the canonical t-exact functor Φ_X from D(Frob(QCoh(X), F_*)) to Frob(D(QCoh(X)), D(F_*)) is an equivalence whenever X is a geometric F_p-scheme, meaning quasi-compact with affine diagonal. On affine schemes, both categories are identified with module spectra over the same E_1-ring, the twisted polynomial ring R[F]^op, using the module-category recognition theorem; the functor matches the compact projective generators. This local identity is then globalized through Zariski descent, since both the source and target presheaves are Zariski sheaves and therefore the equivalence can be checked on affine opens. Because both categories are left-complete, the equivalence

Load-bearing premise

The proof needs the Frobenius pushforward F_* to be an exact functor on quasi-coherent sheaves so that its derived functor is t-exact and the target category inherits a t-structure; this is true for the schemes under consideration because the absolute Frobenius is affine, but the paper leaves it unstated and its introduction even gestures in the opposite direction.

Editorial extensions

If this is right

The derived ∞-category of Frobenius modules satisfies Zariski descent for geometric F_p-schemes, so Frobenius-module cohomology can be computed from affine open covers.
The t-exactness of the equivalence endows D(Frob(QCoh(X), F_*)) with a t-structure whose heart is the ordinary abelian category of Frobenius modules, giving well-behaved homological algebra on singular schemes as well.
Finiteness results that previously required flatness of Frobenius (e.g., for local cohomology and F-modules) can now be phrased and proven in the derived ∞-categorical setting without regularity hypotheses.
The same descent result holds for Cartier modules, so their derived categories are also local objects, useful for duality and finiteness statements.
On affines, both sides become module spectra over R[F]^op, giving a concrete way to compute mapping spectra and Ext groups between Frobenius modules.

Reading between the lines

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

Because the equivalence is t-exact, the heart of D(Frob(QCoh(X), F_*)) is automatically equivalent to the abelian category Frob(QCoh(X), F_*) for any geometric X; the authors do not state this corollary, but it follows directly from Theorem A.
The descent proof uses a nonstandard small Zariski site built from finite disjoint unions of quasi-compact opens, precisely because affine Cech nerves fail without the affine-diagonal assumption; a natural test is whether the descent statements extend to the etale topology or to hypercovers, which the restriction to geometric schemes likely prevents.
The same combination of a limit-preserving lax-equalizer construction, descent, and left-completeness may apply to other endofunctor actions on quasi-coherent sheaves, such as actions of iterated Frobenius or of other affine endomorphisms.
The identification with R[F]^op-module spectra suggests a working definition of Frobenius modules as objects of a stable module category over a twisted polynomial ring, which may simplify concrete computations on singular rings.

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.

Desk Editor's Note

read the letter

First thing to know: the main theorem is very likely correct and extends the authors' earlier Cartier/Frobenius result from regular Noetherian schemes to all geometric F_p-schemes, with Zariski descent as a byproduct. The proof architecture is sensible: affine case via Schwede–Shipley identification with R[F]-module spectra, then Zariski descent, then left-completeness to go from bounded-above to all objects. The descent arguments are genuinely new, and the paper is carefully organized.

The biggest soft spot is one the authors can fix in half an hour: the proof relies on the induced t-structure on Frob(D(QCoh(X)),D(F_*)) existing, which requires D(F_*) to be t-exact. The paper never proves this for geometric schemes, and the introduction actually says that for arbitrary X "this is no longer the case." That statement is wrong: the absolute Frobenius is an affine morphism, so F_* is an exact endofunctor of QCoh(X) and D(F_*) is t-exact. The affine case (Lemma 3.13) gets exactness of F_* from ring-theoretic adjoints, but the global case is left hanging. This is a real gap in the written proof, but the missing lemma is true and the fix is routine. The introduction should also be reworded—what fails without regularity is flatness of F, not exactness of F_*.

The heavier external dependencies—[MW24] for the whole Frob framework, [Pos25] for the Roos axiom, [HM24] for descent—are natural but mean the paper is not self-contained. The authors are upfront about this, and the new theorem genuinely covers cases not in [MW24].

On balance: the paper deserves a serious referee. The main theorem is plausible, the proof strategy is sound, and the gap is small and repairable. I would send it out with a request that the authors state and prove the exactness lemma for F_* globally and correct the introduction. Once that is done, I expect the result to hold up.

Referee Report

3 major / 4 minor

Summary. The paper proves that for a quasi-compact F_p-scheme X with affine diagonal (called geometric), there is a t-exact equivalence of presentable stable ∞-categories D(Frob(QCoh(X), F_*)) ≃ Frob(D(QCoh(X)), D(F_*)). Here Frob(-,-) is the lax-equalizer construction of Frobenius modules from the authors' previous paper [MW24]. The proof has three main steps: (1) an affine reduction, where both sides are identified with module spectra over the same E_1-ring via the Schwede–Shipley theorem; (2) a Zariski-descent argument showing that the bounded-above versions of both sides satisfy descent, so the equivalence is checked on affine opens; (3) a left-completeness argument that promotes the bounded-above equivalence to the full equivalence. The paper also proves Zariski descent for the derived ∞-categories of Frobenius and Cartier modules.

Significance. If the proof is correct, the result is a genuine improvement over [MW24], where the Frobenius-module equivalence required X regular Noetherian. The new statement covers all quasi-compact F_p-schemes with affine diagonal, which is a broad and natural class. The strategy is clear and modular: the affine case is handled by compact projective generators and an endomorphism-ring comparison, the global case by Zariski descent, and the passage from bounded-above objects to all objects by left-completeness. As a byproduct, the paper establishes Zariski descent for derived categories of Frobenius and Cartier modules. The reliance on the companion paper [MW24] for structural facts is substantial but explicit, and the main new theorem is a genuine extension rather than a reformulation. One load-bearing premise, however, is never proved and is even contradicted in the introduction; this must be fixed before the theorem can be considered established.

major comments (3)

[Introduction, p.2; §5, Cor. 5.5] The paper repeatedly uses the fact that D(F_*) is t-exact in order to form the induced t-structure on Frob(D(QCoh(X)), D(F_*)) via [MW24, Prop. 3.3]. This is used in Cor. 5.5 and hence in Thm. 5.7. For geometric X this t-exactness is never proved. Worse, the introduction asserts the opposite: 'If X is arbitrary, then this is no longer the case' (p.2). In fact the absolute Frobenius is affine, so F_*: QCoh(X)→QCoh(X) is exact for every scheme and D(F_*) is t-exact. The assertion is true but the manuscript gives no proof and its wording denies it. This is load-bearing: without exactness of F_*, the target category need not carry the induced t-structure and the left-completeness arguments collapse. Please add a proof (reduction to affine opens, where F_* is restriction of scalars along R→R) and correct the introduction.
[§3, Lemma 3.13] In the affine case, Lemma 3.13 defines a t-structure on Frob(D(Mod_R), D(F_*)) and says this follows from [MW24, Prop. 3.3] because D(F_*) is t-exact 'by definition of the derived functor'. This is missing the essential point that F_* must be exact. The paragraph before Prop. 3.10 asserts that F_* is exact because it has both a left and a right adjoint, but the right adjoint is not written down or referenced in enough detail. Since the global argument in §5 depends on the same property, please give an explicit proof or a precise reference for exactness of F_* both on Mod_R and on QCoh(X) for geometric X.
[§4, Lemma 4.15 and Thm. 4.16] The bounded-above equivalence Φ_X is asserted to be t-exact with the target 'equipped with the induced t-structure from [MW24, Proposition 3.3]' already at the level of global geometric X, not only in the affine case. This is used to identify the sectionwise maps Φ_U with the affine equivalence of Thm. 3.28. The existence of that induced t-structure is precisely the missing exactness of D(F_*) discussed above. The statement is true, but it needs to be proved once, before Lemma 4.15, rather than being inherited from the affine case.

minor comments (4)

[Introduction, p.2] The phrase 'F is flat by Kunz' theorem, and hence F_* is an exact functor' conflates F_* with F^*. Flatness of the Frobenius gives exactness of F^*, whereas F_* is exact without any flatness assumption because the Frobenius is affine. Please rephrase.
[§3, after Lemma 3.3] The statement that F_*: Mod_R → Mod_R is exact 'as it has both a left adjoint and a right adjoint' would be clearer if the right adjoint were exhibited (e.g. Hom_R(R,-) with the appropriate module structure).
[§2, Def. 2.1] The pullback defining End(-) uses '×id_dCat∞' in the top-right spot; this notation is a bit opaque. A sentence explaining the diagram or the identification of the two projections would improve readability.
[§5, Lemma 5.6] The proof says D_{≤n} ≃ D^+_{≤n}; this is correct because D^+ means bounded above, but it would help to emphasize that D_{≤n} is contained in D^+ so the reader does not confuse D^+ with the more common bounded-below convention.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 5.7 is a genuine extension built from an independent affine Schwede–Shipley argument plus Zariski descent; the main weakness is an omitted verification, not a circular reduction, that D(F_*) is t-exact on geometric schemes.

full rationale

The claimed derivation is not circular. The affine case (Theorem 3.28) is proved by identifying both sides with module spectra over the same endomorphism ring: the endomorphism ring equivalence (Proposition 3.27) reduces to Φ^♡ ≃ id, which is proved by a diagram chase from the identification of the hearts (Lemma 3.15) and the naturality of π0; it does not assume the target equivalence. The passage from affine schemes to geometric schemes (Theorem 4.16) uses Zariski descent: the source and target are sheaves (Propositions 4.12 and Lemmas 4.14–4.15), and the morphism is checked on affine opens, where Theorem 3.28 applies. The unbounded theorem (Theorem 5.7) is then a formal left-completeness argument (Lemma 5.6) using Positselski's AB4*_n(ω) theorem [Pos25] and the bounded-above equivalence. The citations to [MW24] supply the definition of Frob(−,−), the induced t-structure on such lax equalizers, and conservativity/limit-preservation of U; these are framework lemmas from the authors' prior paper, not the target result, and the new content — Schwede–Shipley recognition, the R[F]^op endomorphism computation, Zariski descent, and left-completeness — is independent. No fitted parameter is renamed as a prediction, and no uniqueness/ansatz is imported from the authors' own work. One genuine gap should be flagged as a correctness risk rather than circularity: applying [MW24, Proposition 3.3] to define the global induced t-structure on Frob(D(QCoh(X)),D(F_*)) requires D(F_*) to be t-exact, i.e. F_* : QCoh(X) → QCoh(X) exact. The paper does not explicitly prove this for geometric X; the introduction's sentence 'If X is arbitrary, then this is no longer the case' (p. 2) refers to flatness of F, not exactness of F_*, and the missing fact is true because the absolute Frobenius is affine. This is an omitted justification, not a circular reduction.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the exactness of F_* (unstated and contradicted in the intro), on Positselski's AB4*_n(ω) theorem, and on a suite of ∞-categorical results from Lurie, [MW24], and [HM24]. No free parameters or invented entities are introduced; the mathematics is pure and structural.

assumptions (5)

domain assumption Frobenius pushforward F_* is exact on QCoh(X) for any geometric F_p-scheme X, so D(F_*) is t-exact.
Required to apply [MW24, Proposition 3.3] for the induced t-structures on Frob(D(QCoh(X)),D(F_*)) (Corollary 5.5, Theorem 5.7). The paper does not prove this and in the introduction asserts the opposite; the truth follows from the absolute Frobenius being affine, but this is not stated.
domain assumption Positselski's theorem [Pos25, Theorem 1.4]: QCoh(X) satisfies AB4*_n(ω) for geometric X.
Used in Lemma 5.1 to conclude D(QCoh(X)) is left-complete; load-bearing for the upgrade from bounded-above to full equivalence.
standard math Schwede–Shipley theorem [Lur17, Theorem 7.1.2.1] and Gabriel's theorem, in their ∞-categorical forms.
Used in Section 3 to identify affine categories with module spectra over the non-commutative ring R[F]^op.
domain assumption Results of [MW24] on Frob(−,−): existence of left adjoints, conservativity of forgetful functors, preservation of limits, t-structure, and identification of the heart.
The framework is imported wholesale from the same authors' prior preprint; not re-proved here. This is self-citation but not circular in the derivation, as the main theorem is a genuine extension.
standard math [HM24, Proposition A.4.23] on preservation of limit diagrams by derived categories.
Used in Proposition 4.10 to prove Zariski descent of D^+.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The derived $\infty$-category of Frobenius modules." pith.science (2026). https://pith.science/paper/K6V5A6SJ

@misc{pith2026251023267,
  author       = {Pith},
  title        = {Pith review of: The derived $\infty$-category of Frobenius modules},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K6V5A6SJ}},
  note         = {Machine review of arXiv:2510.23267}
}

abstract

We prove that for $X$ a quasi-compact $\mathbb{F}_p$-scheme with affine diagonal (e.g.\ $X$ quasi-compact and separated) there is a t-exact equivalence $\mathcal D(\mathrm{Frob}(\mathrm{QCoh}(X),F_*)) \to \mathrm{Frob}(\mathcal D(\mathrm{QCoh}(X)),\mathcal D(F_*))$ of stable $\infty$-categories. Here, $\mathrm{Frob}(-,-)$ denotes the $\infty$-category of generalized Frobenius modules as introduced in arXiv:2410.17102. This generalizes our result from arXiv:2410.17102, where we proved the above for regular Noetherian $\mathbb{F}_p$-schemes. As a byproduct we prove that the derived $\infty$-category of Frobenius (and Cartier) modules satisfies Zariski descent.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 3 linked inside Pith

[1]

Anderson

Greg W. Anderson. An elementary approach to L -functions mod p . Journal of Number Theory , 80(2):291--303, 2000

2000
[2]

On the uniqueness of infinity-categorical enhancements of triangulated categories

Benjamin Antieau. On the uniqueness of infinity-categorical enhancements of triangulated categories. arXiv preprint arXiv:1812.01526 , 2018

arXiv 2018
[3]

Duality between cartier crystals and perverse F _p -sheaves, and application to generic vanishing, 2025

Jefferson Baudin. Duality between cartier crystals and perverse F _p -sheaves, and application to generic vanishing, 2025

2025
[4]

Cartier modules: finiteness results

Manuel Blickle and Gebhard B \"o ckle. Cartier modules: finiteness results. 2011

2011
[5]

Cartier crystals

Manuel Blickle and Gebhard B \"o ckle. Cartier crystals. arXiv preprint arXiv:1309.1035 , 2013

arXiv 2013
[6]

A. A. Beilinson, J. Bernstein, and P. Deligne. Faisceaux pervers. In Analyse et topologie sur les espaces singuliers (I). CIRM, 6 - 10 juillet 1981 , volume 100 of Ast \'e risque , pages 5--171. Soc. Math. France, Paris, 1982. Luminy conference ; Conference date: 06-07-1981 Through 11-07-1981

1981
[7]

A R iemann- H ilbert correspondence in positive characteristic, 2017

Bhargav Bhatt and Jacob Lurie. A R iemann- H ilbert correspondence in positive characteristic, 2017

2017
[8]

Cohomological Theory of Crystals Over Function Fields

Gebhard B \"o ckle and Richard Pink. Cohomological Theory of Crystals Over Function Fields . EMS tracts in mathematics. European Mathematical Society, 2009

2009

Show all 26 references

[9]

The R iemann- H ilbert correspondence for unit F -crystals

Matthew Emerton and Mark Kisin. The R iemann- H ilbert correspondence for unit F -crystals . Soci \'e t \'e math \'e matique de France, 2004

2004
[10]

Des cat\'egories ab\'eliennes

Pierre Gabriel. Des cat\'egories ab\'eliennes. Bulletin de la Soci\'et\'e Math\'ematique de France , 90:323--448, 1962

1962
[11]

Enriched -categories via non-symmetric -operads

David Gepner and Rune Haugseng. Enriched -categories via non-symmetric -operads. Advances in Mathematics , 279:575–716, July 2015

2015
[12]

Sur quelques points d'alg \`e bre homologique

Alexandre Grothendieck. Sur quelques points d'alg \`e bre homologique. Tohoku Mathematical Journal, Second Series , 9(2):119--183, 1957

1957
[13]

6-functor formalisms and smooth representations

Claudius Heyer and Lucas Mann. 6-functor formalisms and smooth representations. arXiv preprint arXiv:2410.13038 , 2024

2024 arXiv
[14]

Hacon and Zsolt Patakfalvi

Christopher D. Hacon and Zsolt Patakfalvi. Generic vanishing in characteristic p > 0 and the characterization of ordinary abelian varieties. American Journal of Mathematics , 138(4):963--998, 2016

2016
[15]

Local C ohomological D imension in C haracteristic p

Robin Hartshorne and Robert Speiser. Local C ohomological D imension in C haracteristic p . Annals of Mathematics , 105(1):45--79, 1977

1977
[16]

Characterizations of regular local rings of characteristic p

Ernst Kunz. Characterizations of regular local rings of characteristic p . American Journal of Mathematics , 91(3):772--784, 1969

1969
[17]

On the commutation of the test ideal with localization and completion

Gennady Lyubeznik and Karen Smith. On the commutation of the test ideal with localization and completion. Transactions of the American Mathematical Society , 353(8):3149--3180, 2001

2001
[18]

Higher T opos T heory, 2009

Jacob Lurie. Higher T opos T heory, 2009

2009
[19]

Higher A lgebra

Jacob Lurie. Higher A lgebra. Available online at https://www.math.ias.edu/ lurie/papers/HA.pdf, 2017

2017
[20]

Spectral A lgebraic G eometry

Jacob Lurie. Spectral A lgebraic G eometry. Available online at https://www.math.ias.edu/ lurie/papers/sag-rootfile.pdf, 2018

2018
[21]

F -modules: applications to local cohomology and D -modules in characteristic p > 0

Gennady Lyubeznik. F -modules: applications to local cohomology and D -modules in characteristic p > 0 . 1997

1997
[22]

The derived -category of C artier M odules

Klaus Mattis and Timo Weiß. The derived -category of C artier M odules. arXiv preprint arXiv:2410.17102 , 2024

2024
[23]

Frobenius techniques in birational geometry, 2016

Zsolt Patakfalvi. Frobenius techniques in birational geometry, 2016

2016
[24]

Roos axiom holds for quasi-coherent sheaves, 2025

Leonid Positselski. Roos axiom holds for quasi-coherent sheaves, 2025

2025
[25]

F -adjunction

Karl Schwede. F -adjunction. Algebra & Number Theory , 3(8):907--950, 2009

2009
[26]

The stacks project

The Stacks project authors . The stacks project. https://stacks.math.columbia.edu, 2025

2025

Pith tools

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

[1] [1]

Anderson

Greg W. Anderson. An elementary approach to L -functions mod p . Journal of Number Theory , 80(2):291--303, 2000

2000

[2] [2]

On the uniqueness of infinity-categorical enhancements of triangulated categories

Benjamin Antieau. On the uniqueness of infinity-categorical enhancements of triangulated categories. arXiv preprint arXiv:1812.01526 , 2018

arXiv 2018

[3] [3]

Duality between cartier crystals and perverse F _p -sheaves, and application to generic vanishing, 2025

Jefferson Baudin. Duality between cartier crystals and perverse F _p -sheaves, and application to generic vanishing, 2025

2025

[4] [4]

Cartier modules: finiteness results

Manuel Blickle and Gebhard B \"o ckle. Cartier modules: finiteness results. 2011

2011

[5] [5]

Cartier crystals

Manuel Blickle and Gebhard B \"o ckle. Cartier crystals. arXiv preprint arXiv:1309.1035 , 2013

arXiv 2013

[6] [6]

A. A. Beilinson, J. Bernstein, and P. Deligne. Faisceaux pervers. In Analyse et topologie sur les espaces singuliers (I). CIRM, 6 - 10 juillet 1981 , volume 100 of Ast \'e risque , pages 5--171. Soc. Math. France, Paris, 1982. Luminy conference ; Conference date: 06-07-1981 Through 11-07-1981

1981

[7] [7]

A R iemann- H ilbert correspondence in positive characteristic, 2017

Bhargav Bhatt and Jacob Lurie. A R iemann- H ilbert correspondence in positive characteristic, 2017

2017

[8] [8]

Cohomological Theory of Crystals Over Function Fields

Gebhard B \"o ckle and Richard Pink. Cohomological Theory of Crystals Over Function Fields . EMS tracts in mathematics. European Mathematical Society, 2009

2009