Pith. sign in

REVIEW 5 minor 12 references

The geometry of Frobenius on toric varieties

T0 review · 0 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read For a Q-factorial projective toric variety, the Frobenius-trace kernel is ample exactly when the Picard rank is 1.

desk verdict A genuinely new and mostly rigorous toric characterization of ampleness/nefness of the Frobenius-trace kernel; the flagged characteristic-free gap is real but does not threaten the headline theorem. read the letter →

arxiv 2506.02994 v1 pith:RH3FNVAD submitted 2025-06-03 math.AG

classification math.AG MSC 14G1714M2514E3014J4514M17
keywords FrobeniustracekernelsupporttoricvarietiesFanoMoriconePicardrankampleF-signaturepositivecharacteristic
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 establishes a complete geometric dictionary between the Frobenius map of a toric variety and its Mori birational geometry. Its main theorem says that for a Q-factorial projective toric variety in characteristic p>0, the Frobenius-trace kernel E_{X,e} is ample for every e>0 if and only if the Picard rank is 1; it is nef if and only if the variety is a birationally inert Fano variety, meaning every birational extremal contraction is an inert divisorial contraction. The authors identify the object that carries this information, the Frobenius support FS(X), a finite set of divisor classes computed from lattice points in a half-open polytope, and they use it to define cones of F-effective divisors and curves whose interaction with the Mori cone mirrors the types of extremal contractions. This settles the question of when Frobenius positivity encodes global geometry for toric varieties, giving a Frobenius-theoretic counterpart to Mori's characterization of projective space.

What carries the argument

The key object is the Frobenius support $\mathrm{FS}(X)$, together with the cone $\mathrm{Frob}(X)=\langle\mathrm{FS}(X)\rangle_{\mathbb{R}_{\geq 0}}$ of F-effective divisors and its dual cone $\mathrm{FE}(X)=\mathrm{Frob}(X)^\vee$ of F-effective 1-cycles. The machinery works by computing the direct-sum decomposition of $E_{X,e}$ through the splitting formula for Frobenius pushforwards on toric varieties, which reduces the positivity questions to the discrete set of divisor classes $E$ such that $E_{X,e}$ contains $\mathcal{O}_X(E)$; by Corollary 3.6 these are the lattice points of the half-open polytope $Q_X$ spanned by the torus-invariant prime divisors with coefficients in $[0,1)$. Batyrev's primitive relations, which generate the Mori cone $\mathrm{NE}(X)$, are then tested for F-effectiveness: a Mori fibration ray lies in $\mathrm{FE}(X)$, an inert divisorial contraction ray lies in $\mathrm{FE}(X)$ exactly when the contraction is inert, and a small contraction ray does not lie in $\mathrm{FE}(X)$. This trichotomy is what turns the finiteness of $\mathrm{FS}(X)$ into the dichotomy theorems on nefness and ampleness, and it feeds the definition of the ample F-signature $a(X)=\sum_{[E]\in\mathrm{AFS}(X)}\alpha(E)$, whose extremal values detect log Fano structures and homogeneity. An inert divisorial contraction is one whose extremal primitive relation has final coefficient $b_{k+1}=1$; in the smooth case this is exactly a smooth blowup.

What would settle it

Compute the Frobenius support $\mathrm{FS}(X)=Q_X\cap N^1(X)\setminus\{0\}$ for a smooth toric Fano threefold whose fan contains an extremal ray corresponding to a non-maximal-length birational contraction; the paper's Corollary 5.6 predicts an element $[E]\in\mathrm{FS}(X)$ with $E\cdot R<0$, so drawing the lattice points of $Q_X$ and intersecting with the facet of $\mathrm{Nef}(X)$ cut out by $R$ would settle whether $E_{X,e}$ can be nef. If every such $E$ satisfied $E\cdot R\ge 0$, Theorem 5.10 would fail.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the Frobenius-trace kernel $E_{X,e}$ of a $\mathbb{Q}$-factorial projective toric variety $X$ splits as a direct sum of line bundles $\mathcal{O}_X(E)$ indexed by a finite set $\mathrm{FS}(X)\subset N^1(X)$ of numerical divisor classes, the Frobenius support, which is exactly the set of nonzero lattice points of the half-open polytope $Q_X=\langle\pi_1,\ldots,\pi_r\rangle_{[0,1)}\subset\mathrm{Eff}(X)$ (Corollary 3.6). The positivity of $E_{X,e}$ is then a statement about where $\mathrm{FS}(X)$ sits inside the effective cone. Theorem 5.34 proves that $\mathrm{FS}(X)$ is contained in the moving cone exactly when all toric small $\mathbb{Q}$-factorial modifications of $X$ are divisorially inert; $\mathrm{FS}(X)$ is contained in the nef cone exactly when $X$ is a birationally inert Fano variety, in which case inert divisorial contractions reduce $X$ to a variety with $\mathrm{Nef}(X)=\mathrm{Eff}(X)$; and $\mathrm{FS}(X)$ is contained in the ample cone exactly when $\rho(X)=1$, i.e. $X$ is a prime Fano variety. In the smooth case the paper further proves that bigness of $E_{X,e}$ forces $X\simeq\mathbb{P}^d$ (Theorem 4.1) and that nefness is equivalent to $X$ being an extremal Fano variety, with a finite chain of smooth blowups ending at a homogeneous space (Theorem 5.10, Corollary 5.17).

Load-bearing premise

The classification of extremal toric contractions is characteristic-free: every toric Mori fibration is a projective bundle and every birational extremal contraction of maximal length is a smooth blowup, a fact the paper takes from combinatorial arguments and citations rather than proving in detail, and Theorem 5.10 and the smooth case of Theorem 5.34 rely on it.

Editorial extensions

If this is right

  • If a smooth toric variety has nef Frobenius-trace kernel for all $e>0$, then it is an extremal Fano variety and admits a finite chain of smooth blowups ending at a homogeneous space; in particular, its Frobenius support gives a full strong exceptional collection of line bundles on the derived category.
  • If a $\mathbb{Q}$-factorial toric variety has ample Frobenius-trace kernel, it must be a prime Fano toric variety, i.e. of Picard rank 1; these are the fake weighted projective spaces in the singular case.
  • The ample F-signature $a(X)$ is positive exactly when $X$ carries a toric log Fano pair of class index 1, and equals 1 exactly when $\mathrm{Eff}(X)=\mathrm{Nef}(X)$, which for smooth $X$ is equivalent to $X$ being a homogeneous space.
  • Bigness of $E_{X,e}$ on a smooth toric variety forces $X\simeq\mathbb{P}^d$, so within the smooth toric world the Frobenius-trace kernel is ample precisely for projective space, extending Mori's theorem to Frobenius positivity.
  • When the Frobenius support is nef, $|\mathrm{FS}(X)|=|\Sigma(d)|-1$, the number of maximal cones minus one, which measures the length of the full exceptional collection.

Reading between the lines

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

  • The cone of F-effective curves $\mathrm{FE}(X)$ is a new invariant that could be computed algorithmically from the fan: since $\mathrm{FS}(X)$ is just a finite set of lattice points, one can check the conjectural equality $\mathrm{Frob}(X)=\mathrm{Mov}^1(X)$ for toric Fano fourfolds, and the paper's Question 5.37 asks exactly what $\mathrm{FE}(X)\cap\mathrm{NE}(X)$ looks like.
  • The failure of nef primitive relations in the singular case (Remark 4.5 and Example 5.27) suggests that bigness of the Frobenius support alone will not characterize prime Fano varieties outside the smooth case; a more refined invariant, perhaps the full cone $\mathrm{Frob}(X)$ together with the class-index-1 boundary data, would be needed.
  • Because the ample F-signature is defined from asymptotics of direct summand ranks, analogous signatures could be attached to other split Frobenius pushforwards, for instance $F^e_*\mathcal{O}_X(D)$ for fixed $D$, giving a family of numerical invariants that interpolate between $a(X)$ and $n(X)$; the paper computes only the extreme cases.
  • If Question 6.12(b) is true, then $n(X)=1$ would give a purely numerical criterion for $E_{X,e}$ to be nef, and the implication 'ample F-signature equals 1 implies homogeneity' would be a toric analogue of the local F-signature characterization of regularity; testing this on toric Fano threefolds with small Picard rank would be a direct computational check.
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

0 major / 5 minor

Summary. The paper studies the positivity of the Frobenius-trace kernel E_X,e on Q-factorial projective toric varieties in positive characteristic. It introduces the Frobenius support FS(X), the F-effective cone of divisors Frob(X), and the cone of F-effective curves FE(X), and relates them to the Mori geometry of X. The main result (Theorem 5.34) characterizes when FS(X) moves, is nef, or is ample: FS(X) moves iff all small Q-factorial modifications are divisorially inert; FS(X) is nef iff X is a birationally inert Fano variety; and FS(X) is ample iff the Picard rank is 1, which also characterizes when E_X,e is ample. The paper also defines an ample F-signature a(X) and shows a(X)=1 iff X is a homogeneous space in the smooth case. The proofs combine toric combinatorial lemmas with Frobenius splitting results and are largely self-contained.

Significance. If the results hold, this is an important contribution to Frobenius-theoretic birational geometry, providing a toric analog of Mori's theorem and a complete answer to Question 1.1 for toric varieties. The introduction of the Frobenius support and F-effective cones gives new invariants that interact with the Mori cone in a clean way, and the characterization of ampleness by Picard rank 1 is a striking and memorable statement. The paper is also valuable for its connections to Bondal's conjecture, exceptional collections, and the Campana-Peternell conjecture. The proofs are detailed and mostly self-contained, with several characteristic-free combinatorial lemmas that are proved in the text; the authors are honest about points that rely on external references.

minor comments (5)
  1. [Section 2.1.3] The assertion that every toric Mori fibration is a projective bundle in arbitrary characteristic is stated via a brief combinatorial argument and references to [Mon13, Proposition 3.3.8] and [AO02, Theorem 5.1]. The main theorems do not actually rely on this projective-bundle classification: the nefness and ampleness characterizations use only the length computation from Theorem 2.9, namely that a fibration relation has l=k and hence l(R)=d_R+1, together with the inert-divisorial-contraction analysis of Proposition 2.17. Still, since the introduction motivates the term 'extremal Fano' using the projective-bundle description, a direct proof of the toric projective-bundle fact would improve the exposition.
  2. [Theorem 5.10] The proof of Theorem 5.10 is not written out; the text says 'Putting everything together' after Proposition 5.9. Since this is one of the main theorems, an explicit short proof (summarizing the use of Propositions 5.2, 5.9, and Corollary 5.6) would make the paper more accessible and self-contained.
  3. [Proposition 5.24] The notation n := l - (d+1-k) in the statement of Proposition 5.24 gives a negative index when l < d+1-k, although the subsequent sentence says i_l = 0 in that case. The intended meaning is presumably n = max(0, l-(d+1-k)); please clarify this to avoid confusion.
  4. [Corollary 3.6] The inclusion Q_{Z[1/p]} cap N^1(X) is central to the proof of the Frobenius-support formula, but the notation is dense. A short sentence explaining how Lemma 2.20 is applied (with C = Eff(X) and the saturated semigroup from Proposition 2.2) would help the reader.
  5. [Example 6.11] The Macaulay2 computation of the nef F-signature is cited as 'using the computer algebra software Macaulay2 [GS]' without giving the commands or the precise output. Adding a small code snippet or a reference to a reproducible script would strengthen the example.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the main derivation is self-contained; self-citations are supporting, not load-bearing.

full rationale

The derivation is self-contained at the load-bearing points. The Frobenius support FS(X) is computed in Corollary 3.6 from the Thomsen–Achinger splitting theorem (Theorem 3.1) and the elementary lattice lemmas (Lemmas 2.19 and 2.20); this is an external input, not an output of the paper. The nef and ample characterizations (Theorems 5.10 and 5.34) are then derived from the combinatorial primitive-relation calculus (Theorem 2.9, Proposition 2.16) and positivity arguments (Propositions 5.2, 5.5, 5.9, 5.23) that do not presuppose the target statement. Section 2.1.3 states the characteristic-free classification of toric Mori contractions tersely, but the actual steps used later only require the characteristic-free coefficient facts in Theorem 2.9 and the local star-subdivision model of Proposition 2.16; the citations [Mon13] and [AO02] are not the logical engine of the main theorem. The self-citations to [CP21] are used for supporting lemmas such as the projective system (Remark 1.2), the fact that nefness of E_X,e forces Fano, and direct computations in special cases; these are parameter-free results whose assumptions do not include the toric characterization proved here, so they count as real evidence rather than circularity under the review rules. No equation or construction was found that reduces a prediction to a fitted parameter or to a definition.

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

No empirical free parameters were fitted; alpha(E) and m(E;q) are derived from the toric fan. The axioms list the standard background theorems and one characteristic-free assumption on toric Mori contractions. No new physical entities are posited.

assumptions (5)
  • domain assumption X is a Q-factorial projective toric variety over an algebraically closed field of characteristic p>0.
    The entire theorem is stated in this class; Q-factoriality identifies Pic(X) with N^1(X) and enables the combinatorial cone descriptions.
  • standard math Frobenius pushforwards of line bundles on normal toric varieties split into direct sums of line bundles with multiplicities given by the Thomsen-Achinger formula.
    Quoted as Theorem 3.1 from [Ach15, Tho00]; this underpins the computation of the Frobenius support in Corollary 3.6.
  • standard math The effective cone Eff(X) is the simplicial cone generated by invariant divisor classes.
    Proved in Proposition 2.2 using standard toric arguments; used throughout to identify big and Frobenius-supported classes.
  • standard math The Mori cone NE(X) is generated by Batyrev primitive relations.
    Quoted as Theorem 2.6 from [CLS11, Theorem 6.4.11]; used for all extremal contraction arguments.
  • domain assumption For toric varieties in arbitrary characteristic, all Mori fibrations are projective bundles and maximal-length birational extremal contractions are smooth blowups.
    Asserted in Section 2.1.3 with a brief combinatorial justification and citations to [Mon13, Prop 3.3.8] and [AO02, Theorem 5.1]. It is load-bearing for Theorem 5.10 and the nef characterization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The geometry of Frobenius on toric varieties." pith.science (2026). https://pith.science/paper/RH3FNVAD

@misc{pith2026250602994,
  author       = {Pith},
  title        = {Pith review of: The geometry of Frobenius on toric varieties},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RH3FNVAD}},
  note         = {Machine review of arXiv:2506.02994}
}
abstract

We give a geometric description of the positivity of the Frobenius-trace kernel on a $\mathbb{Q}$-factorial projective toric variety. To do so, we define its Frobenius support as well as the notions of $F$-effectiveness for divisors and $1$-cycles. As it turns out, the interaction of the corresponding cone of $F$-effective curves with the Mori cone of curves reflects the type of extremal Mori contractions that the variety can undergo. As a corollary, we obtain that the Frobenius-trace kernel is ample if and only if the Picard rank is $1$.

Figures

Figures reproduced from arXiv: 2506.02994 by the authors.

Figure 1
Figure 1. The N´eron-Severi space of X in Example 5.27 with its nef and pseudo-effective cones as well as Frobenius support. The pseudo-effective cone is generated by π2 and π5. The red rays generate the nef cone. In particular, π1 generates the facet of Nef(X) given by ϕ ∗ (Eff(P 3 )). The green half-open polytope is QX and the purple lattice points are the elements of FS(X). On the other hand, EX,e|XU = EXU ,e, hence this i… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

12 extracted references · 10 canonical work pages

  1. [1]

    The minimal projective bundle dimension and toric $2$-Fano manifolds

    [Ach15] P. Achinger: A characterization of toric varieties in characteristic p, Int. Math. Res. Not. IMRN (2015), no. 16, 6879–6892. 3428948 [A W24] K. Altmann and F. Witt : The structure of exceptional sequences on toric varieties of Picard rank two, Algebr. Comb. 7 (2024), no. 4, 1039–1074. 4804583 [AO02] M. Andreatta and G. Occhetta : Special rays in t...

  2. [3]

    A Series of Modern Surveys in Mathematics [Results in Mathematics and Related Areas

    Folge. A Series of Modern Surveys in Mathematics [Results in Mathematics and Related Areas. 3rd Series. A Series of Modern Surveys in Mathematics], vol. 48, Springer-Verlag, Berlin, 2004, Classical setting: line bundles and linear series. MR2095471 (2005k:14001a) [Laz04b] R. Lazarsfeld: Positivity in algebraic geometry. II , Ergebnisse der Mathematik und ...

  3. [8]

    A Series of Modern Surveys in Mathematics [Results in Mathematics and Related Areas

    Folge. A Series of Modern Surveys in Mathematics [Results in Mathematics and Related Areas. 3rd Series. A Series of Modern Surveys in Mathematics], vol. 49, Springer-Verlag, Berlin, 2004, Positivity for vector bundles, and multiplier ideals. MR2095472 (2005k:14001b) [LMM21] C. Liedtke, G. Martin, and Y. Matsumoto : Linearly Reductive Quotient Singularitie...

  4. [10]

    Mu˜noz, G

    [MOSC+15] R. Mu˜noz, G. Occhetta, L. E. Sol´a Conde, K. W atanabe, and J. a. A. Wi´sniewski: A survey on the Campana-Peternell conjecture , Rend. Istit. Mat. Univ. Trieste 47 (2015), 127–185. 3456582 [RT19] M. Rossi and L. Terracini : A Batyrev type classification of Q-factorial projective toric varieties, Adv. Geom. 19 (2019), no. 4, 433–476. 4015184 [RT...

  5. [30]

    Casagrande: Contractible classes in toric varieties , Math

    4391081 [Cas03] C. Casagrande: Contractible classes in toric varieties , Math. Z. 243 (2003), no. 1, 99–126. 1953051 THE GEOMETRY OF FROBENIUS ON TORIC V ARIETIES 45 [CHMS14] P. Cascini, C. Hacon, M. Musta t ¸˘a, and K. Schwede : On the numerical dimension of pseudo-effective divisors in positive characteristic , Amer. J. Math. 136 (2014), no. 6, 1609–162...

  6. [32]

    3238094 [Wi´ s91] J. a. A. Wi ´sniewski: On contractions of extremal rays of Fano manifolds , J. Reine Angew. Math. 417 (1991), 141–157. 1103910 [Wu24] K.-Y. Wu : Toric varieties with ample tangent bundle , Algebr. Comb. 7 (2024), no. 1, 1–7. 4715529 Centro de Investigaci ´on en Matem´aticas, A.C., Callej ´on Jalisco s/n, 36024 Col. V alen- ciana, Guanaju...

  7. [42]

    Varieties with ample Frobenius-trace kernel

    4801895 [CP21] J. Carvajal-Rojas and Z. Patakfalvi: Varieties with ample Frobenius-trace kernel, arXiv e-prints (2021), arXiv:2110.15035. [CRS23] J. Carvajal-Rojas and A. St¨abler: On the behavior of F -signatures, splitting primes, and test modules under finite covers , J. Pure Appl. Algebra 227 (2023), no. 1, 38 (English), Id/No 107165. [CR22] J. A. Car...

  8. [58]

    Hartshorne: Ample subvarieties of algebraic varieties , Lecture Notes in Mathematics, Vol

    4831691 [Har70] R. Hartshorne: Ample subvarieties of algebraic varieties , Lecture Notes in Mathematics, Vol. 156, Springer-Verlag, Berlin-New York, 1970, Notes written in collaboration with C. Musili. 0282977 [HN13] A. H ¨oring and C. Novelli : Mori contractions of maximal length , Publ. Res. Inst. Math. Sci. 49 (2013), no. 1, 215–228. 3030002 [HL02] C. ...

Show all 12 references
  1. [2005]

    Cartier: Une nouvelle op´ eration sur les formes diff´ erentielles, Comptes rendus de l’Acad´ emie des sciences, Paris 244 (1957), 426–428

    MR2107324 (2005k:14104) [Car57] P. Cartier: Une nouvelle op´ eration sur les formes diff´ erentielles, Comptes rendus de l’Acad´ emie des sciences, Paris 244 (1957), 426–428. 0084497 (18,870b) [CRF24] J. Carvajal-Rojas and A. F ayolle: On tame ramification and centers of F -pu...

  2. [2011]

    [DLMe09] A. Dey, M. Laso ´n, and M. Micha l ek : Derived category of toric varieties with Picard number three, Matematiche (Catania) 64 (2009), no. 2, 99–116. 2800007 [Efi14] A. I. Efimov: Maximal lengths of exceptional collections of line bundles , J. Lond. Math. Soc. (2) 90 ...

  3. [2017]

    Mons ˆores: Toric Birational Geometry and Applications to Lattice Polytopes , arXiv e-prints (2013), arXiv:1307.1449

    [Mon13] D. Mons ˆores: Toric Birational Geometry and Applications to Lattice Polytopes , arXiv e-prints (2013), arXiv:1307.1449. 46 J. CAR V AJAL-ROJAS AND E. A. ¨OZA VCI [Mor79] S. Mori: Projective manifolds with ample tangent bundles , Ann. of Math. (2) 110 (1979), no. 3, 59...

  4. [2021]

    [Tho00] J. F. Thomsen : Frobenius direct images of line bundles on toric varieties , J. Algebra 226 (2000), no. 2, 865–874. 1752764 [Ueh14] H. Uehara: Exceptional collections on toric Fano threefolds and birational geometry , Internat. J. Math. 25 (2014), no. 7, 1450072,

Pith tools

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