Likely intersections in powers of the multiplicative group

Francesco Gallinaro; Gabriel Andreas Dill

REVIEW 2 major objections 3 minor 7 references

Likely intersections in powers of the multiplicative group

T0 review · 2 major / 3 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 For a geometrically non-degenerate subvariety of a multiplicative torus, every translate of a sufficiently large subtorus meets the variety, apart from finitely many exceptional subtori.

desk verdict New finiteness results for likely intersections, but the key uniformity step in Theorem 3.9 has a transfer gap that breaks the main proofs as written. read the letter →

arxiv 2506.07550 v2 pith:XIJHPG4T submitted 2025-06-09 math.NT math.AGmath.LO

Gabriel Andreas Dill , Francesco Gallinaro This is my paper

classification math.NTmath.AGmath.LO MSC 14L1003C9811U0914T90

keywords likelyintersectionsunlikelymultiplicativegroupgeometricalnon-degeneracyrotunditytropicalgeometryequidistributionManin-Mumfordconjecture

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 a finiteness principle for 'likely intersections' in powers of the multiplicative group. If W is an irreducible subvariety of G_m^n that is geometrically non-degenerate, then every translate z·H of a subtorus H with dim H + dim W ≥ n meets W, except when H sits in one of finitely many proper subtori determined by W. The same holds for torsion translates ζ·H, with exceptions lying in finitely many proper algebraic subgroups. This turns a dimension heuristic into a theorem valid in every dimension and explains all failures by a finite list of special subvarieties.

What carries the argument

The load-bearing mechanism is the uniformity theorem (Theorem 3.9): a geometrically non-degenerate W of codimension d admits one radius ε > 0 such that for every d-dimensional real subspace L of the additive group and every z in (C^×)^n, the image of (ℓ, w) ↦ w/(z·exp(ℓ)) contains a ball of radius ε around some point of the unit torus S_1^n. This is obtained by combining a rotundity criterion for the product L × W (openness of the same map, equivalently an amoeba or tropical covering condition) with a study of the tropicalization of W in a non-Archimedean elementary extension of C, followed by a first-order transfer back to the standard reals. The uniform ball is exactly the input needed by the equidistribution theorem for Galois orbits of torsion points, which compares the orbit of a torsion point with all balls of fixed radius.

What would settle it

Find a geometrically non-degenerate W in $G_m^{3}$ and infinitely many distinct two-dimensional subtori H_m, none contained in finitely many fixed proper subtori, each with some translate z_m·H_m disjoint from W — such a family would refute Theorem 1.1 and can be sought computationally for low-degree W. Similarly, infinitely many torsion cosets ζ·H of complement dimension avoiding a non-degenerate W, not lying in finitely many proper algebraic subgroups, would refute Theorem 1.2.

Watch

Extended reading notes

Core claim

The central discovery is that geometrical non-degeneracy of W forces its intersections with cosets of subtori of dimension at least codim W to be ubiquitous. Precisely, Theorem 1.1 gives a finite set of proper subtori H_i such that for every subtorus H with dim H + dim W ≥ n, either every complex translate of H meets W or H is contained in some H_i. Theorem 1.2 gives the analogous statement for torsion translates, with the exceptions being finitely many proper algebraic subgroups. The engine behind both is a uniformity result: for a codimension-d non-degenerate W there is a fixed ε > 0 such that for every d-dimensional real linear subspace L and every z, the quotient $z^{{-1}}$W(C)/exp(L(C)) contains a Euclidean ball of radius ε centered on the unit torus; equidistribution of Galois orbits of torsion points then forces the intersections.

Load-bearing premise

The whole argument hangs on the uniformity statement that a single fixed radius works for every subspace direction and every translate; that statement is proved by transferring a property from a non-Archimedean extension of the complex numbers back to the standard reals.

Editorial extensions

If this is right

In characteristic zero, the only empty likely intersections with general translates are certified by finitely many proper subtori; the only empty torsion ones by finitely many proper algebraic subgroups.
A new proof of the Manin-Mumford conjecture for tori follows: the torsion points of a subvariety are contained in finitely many torsion cosets (Proposition 5.2).
For an irreducible curve not contained in a coset, all but finitely many cosets of codimension-one subtori meet the curve, and the exceptions are finitely many explicit cosets (Corollary 1.3).
For a non-degenerate W of dimension d, the fraction of integer matrices of size d × n with entries bounded by N for which the associated monomial map is surjective on W tends to 1 as N → ∞ (Theorem 1.4).
Theorems 1.1 and 1.2 extend verbatim to every algebraically closed field of characteristic 0 (Theorem 5.1).

Reading between the lines

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

A likely follow-up is an effective version: the finite exceptional lists in Theorems 1.1 and 1.2 arise from a Gröbner basis and a finite set of tropical polyhedra, so explicit bounds on their size and complexity should be extractable from the proof.
The uniformity radius idea may transfer to other uniformization settings, where an analogous 'fixed ball' statement would yield likely-intersection finiteness for the corresponding special subvarieties.
The paper's hypersurface example shows the torsion statement is optimal, but also suggests a testable quantitative question: how the exceptional algebraic subgroups grow with the height or degree of W.
Theorem 1.4 suggests the probabilistic interpretation is robust; one could test whether the density-one conclusion persists for merely rotund, rather than geometrically non-degenerate, W.

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

The main results are new and natural: Theorems 1.1 and 1.2 turn known density statements for likely intersections in G_m^n into full finiteness under geometric non-degeneracy. The strategy—uniform rotundity via tropicalization, then Bilu equidistribution—is attractive, and the paper is carefully written, with honest discussion of sharpness and limitations. The citation practice looks reasonable.

The problem is in Theorem 3.9. The proof establishes, for each fixed standard L and z, a positive real epsilon and a ball B(s1,epsilon) contained in z^{-1}W/exp(L). The final paragraph then asserts that the first-order sentence "exists epsilon>0 forall L,z exists s1 B(s1,epsilon) subset ..." holds in the nonstandard model R because any positive infinitesimal is a witness, and transfers this to Rexp,sin. That inference is not justified. Pointwise existence of positive radii for each standard L,z does not give a single positive radius, even a nonstandard one, that works uniformly for all internal L,z. A simple definable family—P(x,epsilon): epsilon<|x| on (0,1)—has a positive radius for each x but no uniform positive radius. The proof would need a compactness or continuity argument to get joint uniformity in L and z before the transfer is legitimate.

This matters because the uniform epsilon is exactly what feeds Lemma 4.2 and then Theorems 1.1 and 1.2. So as written, the main theorems are not proven. The rest of the paper is solid and the gap looks fixable, but it is load-bearing.

The reader's take was more optimistic than the text supports. I agree on novelty and importance; the soundness score should come down until the transfer in Theorem 3.9 is repaired. The paper deserves a serious referee—the ideas are good and the writing is honest—but the referee should be asked to focus on that transfer.

Recommendation: send to peer review, with a referee familiar with nonstandard methods and tropical geometry. Expect a major revision.

Referee Report

2 major / 3 minor

Summary. The paper proves two finiteness theorems for geometrically non-degenerate subvarieties W of G_m^n over C. Theorem 1.1 shows that for every subtorus H with dim H + dim W >= n, either every translate z·H meets W, or H is contained in one of finitely many proper subtori depending only on W. Theorem 1.2 gives the analogous statement for torsion translates: either every torsion translate zeta·H meets W, or zeta·H is contained in one of finitely many proper algebraic subgroups. The proofs combine tropical geometry, rotundity characterizations from prior work, a uniformity statement (Theorem 3.9) proved via an elementary extension of the reals with exponentials and sine, and Bilu's equidistribution theorem. The paper also derives applications, including a new proof of the Manin-Mumford conjecture for tori, a strengthening for curves, and a probabilistic density statement for monomial maps.

Significance. If the central uniformity statement holds, these are strong and appealing results: they show that for geometrically non-degenerate W, failures of likely intersections are certified by a finite list of proper subtori (respectively algebraic subgroups), uniformly in the translate. The paper is carefully written, states its hypotheses explicitly, and honestly acknowledges limitations, including failure over finite fields and non-inheritance of geometrical non-degeneracy. The tropical/rotundity equivalences, the equidistribution argument, and the applications to Manin-Mumford and to monomial maps are substantial contributions. The main proof is intricate and the nonstandard transfer is the most fragile point; it needs to be made rigorous before the results can be considered established.

major comments (2)

[Section 3.2, Theorem 3.9 (final paragraph)] The final transfer step is not justified. The proof establishes, for each standard L in GR(d,n) and each z in (C×)^n, the existence of a standard real radius epsilon(L,z)>0 and s1 such that B(s1,epsilon(L,z)) is contained in z^{-1}W(C)/exp(L(C)). From this pointwise statement the text infers that the sentence exists epsilon>0 forall L forall z exists s1 (B(s1,epsilon) subset z^{-1}W/exp(L)) holds in the nonstandard model, 'because any positive infinitesimal element is a witness for it'. This inference is invalid: in an elementary extension one can have forall x exists epsilon>0 P(x,epsilon) without exists epsilon>0 forall x P(x,epsilon) (e.g., P(n,epsilon) = 'epsilon < 1/n' on the natural numbers), and a fixed infinitesimal eta need not be <= epsilon(L,z) for every internal pair (L,z), since the transferred pointwise statement only gives some positive radius for each pair. The radii produced earlier in the proof depend on z through the multiplier c = res(s(alpha)z'^{-1}y) and can be arbitrarily small; no lower bound uniform in z is shown. Because the constant epsilon of Theorem 3.9 is the input to Lemma 4.2 and Lemma 4.3, and hence to Theorems 1.1 and 1.2, this is a load-bearing gap. The authors need either a direct uniformity argument (e.g., a compactness/continuity argument over the parameter space) or a different transfer statement that does not move a pointwise radius to a uniform one.
[Section 3.2, proof of Theorem 3.9, after Eq. (3.2)] In the same proof, the step 'By Proposition 3.5(8), there is a non-empty open subset U of S1^n(C) such that U is contained in res(s(alpha)z'^{-1}y) · W_tau(C)/exp(res(L)(C))' needs justification: Proposition 3.5(8) only yields a ball around a point with the same moduli as an arbitrary input, not an open subset of the unit torus; moreover the radius of the ball obtained from openness of delta on W°_tau is attached to a point s0 that depends on z, so the subsequent epsilon0 is not shown to be uniform. This is part of the same pointwise/uniformity gap and should be addressed in the revision.

minor comments (3)

[Sections 2.3–3.2] The two typographically similar symbols R and R are used for the standard reals and the elementary extension; in the final paragraph of Theorem 3.9 this distinction is load-bearing and should be typeset more distinctly (for example, using *R for the extension).
[Introduction, paragraph after Theorem 1.2] The hypersurface example W0 defined by x+2y+z=1 is said to have trivial stabilizer and therefore to be geometrically non-degenerate; this uses the hypersurface criterion stated earlier in the section, but a short explicit hint would improve readability.
[References] The reference [G23] appears to be missing the author's full name; please check the bibliographic entry.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main theorems are genuinely derived from a new uniformity theorem and prior published rotundity results, not from definitions or fitted parameters.

full rationale

The paper derives Theorems 1.1 and 1.2 from a new uniformity statement, Theorem 3.9. Geometrical non-degeneracy is defined by dimension inequalities for quotient maps, while the conclusions assert that all but finitely many subtori have every translate meeting W; the finite exceptional families are not defined to be the set of failures but are produced by Lemma 4.3 from Bilu equidistribution together with the uniform ball property, so the result is not true by construction. The proof of Theorem 3.9 uses the rotundity characterizations of Proposition 3.5, which partly cite [Gal23a] (by the second author) and [Kir19]. These are published, parameter-free results whose assumptions do not include the geometrical non-degeneracy conclusion or the uniformity theorem, so they function as independent support rather than as a self-citation chain. The most delicate step is the final transfer in Theorem 3.9: the proof establishes a pointwise standard radius for each fixed L and z and then invokes an infinitesimal witness to obtain a uniform epsilon. This inference may be a correctness gap, since pointwise positive radii over a family do not automatically yield a uniform positive radius without an additional compactness or continuity argument, but it is not circular because the claimed uniform conclusion is not an input to the pointwise proof. No fitted constants are renamed as predictions; the exceptional subgroups in Lemmas 4.2-4.3 depend only on n, K, and epsilon, not on W. Thus no circular reduction is present.

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

The paper introduces no free parameters and no new entities. Its main results rest on standard background theorems in tropical geometry, model theory of valued fields, and equidistribution of torsion points, together with the theorem's own hypothesis that W is geometrically non-degenerate. The most paper-specific ingredient is the construction of a large non-Archimedean field extension of C and the model-theoretic transfer used in Theorem 3.9; these are standard but load-bearing.

assumptions (5)

standard math Bilu's equidistribution theorem for torsion points of algebraic tori (Theorem 4.1)
Used in Lemma 4.2 to make Galois orbits of high-order torsion points hit every fixed-radius ball around S1^n; this is the equidistribution engine.
standard math Tropical geometry results of Maclagan and Sturmfels: val(W(K)) is a polyhedral complex, initial varieties have dimension dim W, initial varieties are stabilized by J_tau, and the star of a tropicalization is a tropicalization of the initial variety
Section 2.2, Theorems 2.6 and 2.7, Propositions 2.11 and 2.14; used to relate rotundity to the coverage condition (6) and to prove Proposition 3.8.
standard math Completeness, quantifier elimination, and model completeness for algebraically closed non-trivially valued fields of residue characteristic 0
Used in Section 2.3 to transfer statements between C, C0, and C, and at the end of Theorem 3.9 to transfer a uniform epsilon back to the standard structure Rexp,sin.
standard math Existence of a proper elementary extension R of Rexp,sin with the transfer principle
Section 2.3; needed for the nonstandard proof of uniformity in Theorem 3.9.
standard math Complex analytic facts: open mapping theorem, upper semicontinuity of fiber dimension, and the fact that no open subset of the unit torus lies in a countable union of lower-dimensional analytic sets
Used in Lemma 3.3, Proposition 3.5, and Theorem 3.9 to convert local open maps into balls.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Likely intersections in powers of the multiplicative group." pith.science (2026). https://pith.science/paper/XIJHPG4T

@misc{pith2026250607550,
  author       = {Pith},
  title        = {Pith review of: Likely intersections in powers of the multiplicative group},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XIJHPG4T}},
  note         = {Machine review of arXiv:2506.07550}
}

abstract

We derive two finiteness properties as consequences of the geometrical non-degeneracy of an algebraic subvariety $W$ of a power of the multiplicative group, concerning the intersections of $W$ with translates of a subtorus $H$ of dimension greater than or equal to the codimension of $W$. The first one is that every translate of $H$ intersects $W$, unless $H$ is contained in one of finitely many proper subtori depending only on $W$. The second one is that every translate of $H$ by a torsion point intersects $W$, unless the translate is contained in one of finitely many proper algebraic subgroups, again depending only on $W$. We use methods from tropical geometry and equidistribution, as well as some very mild model theory.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

7 extracted references · 7 canonical work pages

[4]

Algebraic varieties and auto- morphic functions

arXiv: 2211.10592. [EZ25] Sebastian Eterovi´ c and Roy Zhao. “Algebraic varieties and auto- morphic functions”. In: Int. Math. Res. Not. 2025.4 (2025). doi: 10.1093/imrn/rnaf029. [G23] Asvin G. Unlikely and just likely intersections for high dimen- sional families of elliptic curves . 2023. arXiv: 2203.16420. [Gal23a] Francesco Gallinaro. “Exponential sum...

work page arXiv 2025
[236]

Degenerations of amoebae and Berkovich spaces

doi: doi:10.1515/CRELLE.2006.066. [Hod97] Wilfrid Hodges. A shorter model theory . Cambridge University Press, 1997. [Isa09] I. Martin Isaacs. Algebra: a graduate course . Vol. 100. Graduate Studies in Mathematics. Reprint of the 1994 original. American Mathematical Society, Providence, RI, 2009, pp. xii+516. doi: 10.1090/gsm/100. [Jon16] Mattias Jonsson....

work page doi:10.1515/crelle.2006.066 2016
[454]

Exceptional isogenies between reductions of pairs of elliptic curves

doi: 10.1112/jlms.12014. [Cha18] Fran¸ cois Charles. “Exceptional isogenies between reductions of pairs of elliptic curves”. In:Duke Math. J. 167.11 (2018), pp. 2039–

work page doi:10.1112/jlms.12014 2018
[2023]

A geometric approach to some systems of exponential equations

arXiv: 2312.09974. [AKM23] Vahagn Aslanyan, Jonathan Kirby, and Vincenzo Mantova. “A geometric approach to some systems of exponential equations”. In: Int. Math. Res. Not. 2023 (2023), pp. 4046–4081. doi: 10. 1093/imrn/rnab340. [BIR08] Matthew Baker, Su-ion Ih, and Robert Rumely. “A finiteness property of torsion points”. In:Algebra Number Theory 2.2 (200...

work page arXiv 2023
[2072]

Solutions of equa- tions involving the modular j function

doi: 10.1215/00127094-2018-0011. [Chi89] Evgenii Mikhailovich Chirka. Complex analytic sets. Vol. 46. Math- ematics and its Applications (Soviet Series). Translated from the Russian by R. A. M. Hoksbergen. Kluwer Academic Publishers Group, Dordrecht, 1989, pp. xx+372. doi: 10.1007/978- 94- 009-2366-9. [Dri98] Lou van den Dries. Tame topology and o-minimal...

work page doi:10.1215/00127094-2018-0011 2021
[2748]

Integral division points on curves

doi: 10.1112/S0010437X21007673. [GI13] David Grant and Su-ion Ih. “Integral division points on curves”. In: Compos. Math. 149.12 (2013), pp. 2011–2035. doi: 10.1112/ S0010437X13007318. [GR84] Hans Grauert and Reinhold Remmert. Coherent analytic sheaves. Vol. 265. Grundlehren der mathematischen Wissenschaften. Springer Science & Business Media, 1984. doi: ...

work page doi:10.1112/s0010437x21007673 2013
[5366]

A common generalization of the conjectures of Andr´ e-Oort, Manin-Mumford, and Mordell-Lang

doi: 10.1093/imrn/rnq248. [Pin05] Richard Pink. “A common generalization of the conjectures of Andr´ e-Oort, Manin-Mumford, and Mordell-Lang”. https : / / people.math.ethz.ch/~pink/ftp/AOMMML.pdf. 2005. [Ran81] Ziv Ran. “On subvarieties of abelian varieties”. In: Invent. Math. 62.3 (1980/1981), pp. 459–479. doi: 10.1007/BF01394255. [Rob56] Abraham Robinso...

work page doi:10.1093/imrn/rnq248 2020

Pith tools

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

[1] [4]

Algebraic varieties and auto- morphic functions

arXiv: 2211.10592. [EZ25] Sebastian Eterovi´ c and Roy Zhao. “Algebraic varieties and auto- morphic functions”. In: Int. Math. Res. Not. 2025.4 (2025). doi: 10.1093/imrn/rnaf029. [G23] Asvin G. Unlikely and just likely intersections for high dimen- sional families of elliptic curves . 2023. arXiv: 2203.16420. [Gal23a] Francesco Gallinaro. “Exponential sum...

work page arXiv 2025

[2] [236]

Degenerations of amoebae and Berkovich spaces

doi: doi:10.1515/CRELLE.2006.066. [Hod97] Wilfrid Hodges. A shorter model theory . Cambridge University Press, 1997. [Isa09] I. Martin Isaacs. Algebra: a graduate course . Vol. 100. Graduate Studies in Mathematics. Reprint of the 1994 original. American Mathematical Society, Providence, RI, 2009, pp. xii+516. doi: 10.1090/gsm/100. [Jon16] Mattias Jonsson....

work page doi:10.1515/crelle.2006.066 2016

[3] [454]

Exceptional isogenies between reductions of pairs of elliptic curves

doi: 10.1112/jlms.12014. [Cha18] Fran¸ cois Charles. “Exceptional isogenies between reductions of pairs of elliptic curves”. In:Duke Math. J. 167.11 (2018), pp. 2039–

work page doi:10.1112/jlms.12014 2018

[4] [2023]

A geometric approach to some systems of exponential equations

arXiv: 2312.09974. [AKM23] Vahagn Aslanyan, Jonathan Kirby, and Vincenzo Mantova. “A geometric approach to some systems of exponential equations”. In: Int. Math. Res. Not. 2023 (2023), pp. 4046–4081. doi: 10. 1093/imrn/rnab340. [BIR08] Matthew Baker, Su-ion Ih, and Robert Rumely. “A finiteness property of torsion points”. In:Algebra Number Theory 2.2 (200...

work page arXiv 2023

[5] [2072]

Solutions of equa- tions involving the modular j function

doi: 10.1215/00127094-2018-0011. [Chi89] Evgenii Mikhailovich Chirka. Complex analytic sets. Vol. 46. Math- ematics and its Applications (Soviet Series). Translated from the Russian by R. A. M. Hoksbergen. Kluwer Academic Publishers Group, Dordrecht, 1989, pp. xx+372. doi: 10.1007/978- 94- 009-2366-9. [Dri98] Lou van den Dries. Tame topology and o-minimal...

work page doi:10.1215/00127094-2018-0011 2021

[6] [2748]

Integral division points on curves

doi: 10.1112/S0010437X21007673. [GI13] David Grant and Su-ion Ih. “Integral division points on curves”. In: Compos. Math. 149.12 (2013), pp. 2011–2035. doi: 10.1112/ S0010437X13007318. [GR84] Hans Grauert and Reinhold Remmert. Coherent analytic sheaves. Vol. 265. Grundlehren der mathematischen Wissenschaften. Springer Science & Business Media, 1984. doi: ...

work page doi:10.1112/s0010437x21007673 2013

[7] [5366]

A common generalization of the conjectures of Andr´ e-Oort, Manin-Mumford, and Mordell-Lang

doi: 10.1093/imrn/rnq248. [Pin05] Richard Pink. “A common generalization of the conjectures of Andr´ e-Oort, Manin-Mumford, and Mordell-Lang”. https : / / people.math.ethz.ch/~pink/ftp/AOMMML.pdf. 2005. [Ran81] Ziv Ran. “On subvarieties of abelian varieties”. In: Invent. Math. 62.3 (1980/1981), pp. 459–479. doi: 10.1007/BF01394255. [Rob56] Abraham Robinso...

work page doi:10.1093/imrn/rnq248 2020