Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Towards a p-adic nowhere density conjecture of Hecke Orbits

T0 review · 3 major / 4 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 The paper establishes that around a mod $p$ point of a Hodge-type Shimura variety, the points of the formal neighborhood with large $p$-adic monodromy form an open dense set, giving a local version of the proposed $p$-adic nowhere density…

desk verdict A plausible and genuinely new formal density result for Hodge-type Shimura varieties, but the proof as written does not establish the stated theorem. read the letter →

arxiv 2506.06932 v1 pith:CQENLUN5 submitted 2025-06-07 math.NT math.AG

Yu Fu This is my paper
classification math.NTmath.AG MSC 11G1814L0514G3511F80
keywords p-adicmonodromyHeckeorbitsShimuravarietiesHodgetypep-divisiblegroupsBreuilmodulesformalneighborhoodsnowheredensity
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 proposes a conjecture: in the $p$-adic topology on a Shimura variety, the Hecke orbit of any positive-dimensional subvariety should be nowhere dense, rather than dense as it is in the archimedean topology. The main theorem proves a local version of this. Fix a mod $p$ point whose Newton and Hodge polygons only meet at their endpoints, let $K$ be a totally ramified finite extension of the completed maximal unramified extension of the reflex field, and look at the $K$-integral points in the formal neighborhood of that point whose special fiber is the same $p$-divisible group. Within that set, the points whose $p$-adic Galois representation has large monodromy are open and dense. This is the local density input that the full conjecture needs, and it comes with the corollary that prime-to-$p$ Hecke orbits of such points are $p$-adically nowhere dense.

What carries the argument

The argument runs on a congruence-to-monodromy mechanism for $p$-divisible groups. Lemma 2.5, obtained from the classification of finite flat group schemes by Breuil modules, says that if two $p$-divisible groups agree modulo $p^{n+1}$, their $p^n$-torsion is isomorphic as group schemes. Proposition 2.7 then propagates large monodromy: if one group has Galois image containing the congruence subgroup $H_n = G^{\mathrm{der}}(\mathbb{Z}_p) \cap K_n$ with $K_n = \{ M \in \mathrm{Sp}_{2g}(\mathbb{Z}_p) : M \equiv I \bmod p^n \}$, any group whose $p^{n+1}$-torsion is isomorphic has image containing some $H_m$; the proof uses the $p$-power map on congruence quotients, whose surjectivity it asserts. This makes the large-monodromy locus open. For density, the paper uses Rapoport-Zink spaces of Hodge type (formal moduli of deformations of a $p$-divisible group with crystalline Tate tensors), the rigid analytic period map to a $p$-adic flag variety, the connectedness of the admissible period domain, and a lemma that generic points of the period domain are $p$-adically close to any given point; generic points are already known from the theory of $p$-adic period domains to have large monodromy. The uniformization theorem identifies formal neighborhoods in the Shimura variety with quotients of these Rapoport-Zink spaces, which is why the local statement transfers to $S_{K,x}$.

What would settle it

Exhibit two $p$-divisible groups over $\mathcal{O}_K$ whose reductions modulo $p^{n+1}$ are isomorphic as group schemes, where the first has Galois image containing a congruence subgroup of $G^{\mathrm{der}}(\mathbb{Q}_p)$ and the second has Galois image containing no such subgroup; such a pair would falsify Proposition 2.7 and remove the openness assertion of Theorem 1.5.

Watch

Extended reading notes

Core claim

Let $x$ be an $\mathbb{F}_p$-point of a Hodge-type Shimura variety and let $(G,[b],\mu)$ be the associated local Hodge-Shimura datum, with $A_x[p^\infty]$ the $p$-divisible group of the abelian variety at $x$. Assume the Newton and Hodge polygons of $[b]$ and $\mu$ do not touch outside their endpoints (HN-irreducibility). For a totally ramified finite extension $K$ of the completed maximal unramified extension of the reflex field, let $S_{K,x}$ be the set of $\mathcal{O}_K$-points of the formal neighborhood of $x$ whose special fiber is isomorphic to $A_x[p^\infty]$. The paper proves that the subset $S_{K,x}^{\max}$ of points whose Galois representation $\rho_y$ has image containing an open subgroup of $G^{\mathrm{der}}$ is $p$-adically open and dense in $S_{K,x}$. It reaches this by a different route than the earlier PEL case: openness comes from a congruence criterion for monodromy proved with the classification of finite flat groups, and density comes from producing generic points, which necessarily have large monodromy, in every $p$-adic neighborhood of any point. Theorem 1.5 then yields Corollary 1.6: the prime-to-$p$ Hecke orbit of a point with good reduction and HN-irreducible local datum is nowhere dense in the $p$-adic topology on $\mathrm{Sh}(\mathbb{C}_p)$.

Load-bearing premise

The proof rests on the propagation lemma that two $p$-divisible groups with the same reduction modulo $p^{n+1}$ have the same large-monodromy property, and that lemma relies on two technical steps the paper asserts rather than fully proves: surjectivity of the $p$-power map on congruence quotients and the claim that Breuil-module invariants are determined by reduction modulo $p^{n+1}$.

Editorial extensions

If this is right

  • If Theorem 1.5 holds, the prime-to-$p$ Hecke orbit of any point of $\mathrm{Sh}(\mathbb{C}_p)$ with good reduction and HN-irreducible local datum is $p$-adically nowhere dense.
  • The formal-local statement is a first step toward the full conjecture: one would need the same open-density statement for all $\mathbb{C}_p$-deformations, not only those over fixed finite extensions of $\breve{E}$.
  • If the full conjecture holds, the set of CM points in $\mathrm{Sh}(\mathbb{C}_p)$ is closed in the rigid analytic topology, and the Hecke orbit of the Torelli locus is small enough to yield new statements about which abelian varieties are isogenous to Jacobians.
  • The proof strategy gives an alternative route to the earlier PEL-type result, replacing $p$-adic Lefschetz arguments with the classification of finite flat groups and Rapoport-Zink uniformization.

Reading between the lines

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

  • If the propagation lemma can be strengthened to allow bounded wild ramification of the base field, the formal-local statement may extend from finite extensions of $\breve{E}$ to all $\mathbb{C}_p$-deformations, which is exactly the gap the paper identifies.
  • The same congruence-propagation mechanism might prove analogous open-density statements for other classes of $p$-divisible groups or local Shimura varieties, since the input is the classification of finite flat groups by filtered modules rather than a special feature of abelian varieties.
  • A testable consequence beyond the paper's scope: if the Hecke orbit of the Torelli locus is $p$-adically nowhere dense, then generic abelian varieties over $\mathbb{Q}$ are not quotients of Jacobians; the paper notes this direction without proving it.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes Conjecture 1.3, a p-adic nowhere-density statement for Hecke orbits of positive-dimensional subvarieties of Hodge-type Shimura varieties. Its main theorem (Theorem 1.5) asserts that, under an HN-irreducibility condition, the locus S_{K,x}^{max} of O_K-points in the formal neighborhood of a mod p point x whose p-adic Galois representation has large monodromy is p-adically open dense. The proof has two parts: Theorem 2.1 establishes openness using Breuil's classification of finite flat group schemes and a congruence-subgroup propagation lemma (Proposition 2.7); Proposition 3.7 establishes density using p-adic uniformization, period domains, and generic points in F^a. Corollary 1.6 derives p-adic nowhere density of prime-to-p Hecke orbits.

Significance. If the main theorem were proved as stated, it would be a significant step toward the proposed conjecture and would generalize Maulik-Poonen's formal density result to Hodge-type Shimura varieties by a genuinely different method (Breuil modules plus period-domain geometry). The paper is readable and honest about limitations, explicitly flagging the obstruction to extending the result to C_p-points. It also imports its main external inputs from Chen, Gleason-Lim-Xu, and Gleason-Lourenco rather than relying on circular self-support. However, the current proofs of both the openness and density halves contain load-bearing gaps, so the advertised Theorem 1.5 is conditional on work that is not supplied in the manuscript.

major comments (3)
  1. [Section 3.6, Proposition 3.7 (and Theorem 1.5)] The density half of Theorem 1.5 is not proved for the original field K. In the proof of Proposition 3.7, the generic point z is produced in F^a(K) by Lemma 3.6, but the fiber \bar\pi^{-1}(z) is only shown to admit a point after replacing K by a finite extension K': the proof explicitly says 'any connected component of it is a finite extension K' of K. After replacing K with K', we obtain a K-point y0'. Thus the constructed large-monodromy point lies in S_{K',x}, not in S_{K,x}. Since 'large monodromy' in Definition 1.4 is a property of the Gal(\bar K/K)-representation, a K'-point does not produce an element of S_{K,x}^{max}; and the text after Corollary 1.6 acknowledges that passing to wild ramified extensions can shrink monodromy, so there is no descent argument. Consequently Proposition 3.7 establishes only density after base change, and Theorem 1.5, which drops this caveat, is strictly stronger than what is proved.
  2. [Section 2.3, Proposition 2.7 (Eq. (2.9))] The openness half rests on the claim that the p-power map H_m/H_{m+1} \to H_{m+1}/H_{m+2} is surjective for some m \ge n. The proof computes the analogous map for GL_{2g} and then says the general case follows from naturality of the p-adic exponential map. This is not a complete argument: the exponential/logarithm is only a local bijection between a neighborhood of the identity and a Lie algebra lattice, and it does not directly identify the graded pieces H_m/H_{m+1} and H_{m+1}/H_{m+2} for an arbitrary reductive Z_p-model G^{der}. A precise congruence-subgroup calculation, or a reference, is needed. Without it, the induction step in Proposition 2.7 and hence the openness conclusion of Theorem 2.1 are unsupported.
  3. [Section 2.2, Lemma 2.5 and its use in Proposition 2.7] Lemma 2.5, which converts p-adic congruence of lifts into isomorphism of p^{n+1}-torsion, is proved only by a sketch. The claim that the Breuil-module invariants (M, Fil^1 M, \phi_1) are determined by the reduction modulo p^{n+1} is asserted rather than derived: the diagram chase for the divided Frobenius \phi_1 assumes the embeddings (2.5)-(2.6) and the required compatibility with the isomorphism \psi, which are essentially the statements needing proof. Since Theorem 2.1 applies Lemma 2.5 with level n+1, this gap is load-bearing for the openness half.
minor comments (4)
  1. [Throughout] There are numerous typographical artifacts (for example, 'thep-adic' in the abstract, 'OK' for O_K, and missing subscripts in 'Ocris'), and the reference formatting is inconsistent; a careful proofread is needed.
  2. [Section 3.6, proof of Proposition 3.7] The final sentence says the constructed point 'maps to a OK point z \in V_x'; it should be an O_{K'}-point, and the reuse of z for both the generic period and the image point is confusing.
  3. [Section 2.3, Proposition 2.7] The definition of L_i as 'H_i/H_{i+1} mod p^{i+1}' conflates a quotient group with a set of matrices; these should be defined as congruence subgroups in G^{der}(Z_p), not as quotients.
  4. [Section 3.6, Lemma 3.6] Lemma 3.6 reuses the symbol K0 from Section 2.1 without redefinition; in Section 3 the reader must infer that K0 denotes the completed maximal unramified extension of Q_p.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the proof imports external theorems and develops an independent congruence argument; the main gap (field-extension density) is an incompleteness, not a circularity.

full rationale

The paper's derivation chain is not circular. Theorem 1.5 is presented as following from Theorem 2.1 and Proposition 3.7. Theorem 2.1 is an internal congruence-propagation statement: if one O_K-lift has Galois image containing H_n, then any O_K-lift congruent to it modulo p^{n+2} also has large monodromy. Its proof uses Breuil's classification of finite flat group schemes (external, [Bre00]) and an internal diagram chase (Lemma 2.5) plus the p-power map on congruence subgroups (Proposition 2.7). Even where these arguments are asserted rather than fully proved, they do not presuppose the target theorem; they are independent mathematical inputs. Proposition 3.7, the density half, relies on Chen's and Gleason-Lim-Xu's theorem that generic points of the period domain have large monodromy ([Che14, Theorem C], [GLX23, Theorem 1.14]), on connectedness of period domains ([GL22, Theorem 1.1]), and on Colmez-Fontaine. None of these are works of the present author, and none are fitted to the conclusion; they are cited external theorems. The paper itself flags the limitation that the strategy cannot cover all C_p-deformations and that Proposition 3.7 only gives density 'after possibly replacing K with a finite extension'. Since the proof only produces large-monodromy points over K' (a finite extension of K), the asserted density over the original K in Theorem 1.5 is not established. This is a genuine mathematical gap in the claimed derivation chain, but it is a gap, not a circular reduction: the conclusion is not being used as an input, and no parameter is fitted so as to force the result. There are no self-citations carrying load, no uniqueness theorem imported from the authors' own prior work, and no empirical quantity being renamed as a prediction. The correct circularity reading is therefore a clean non-finding.

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

The proof rests on several major external theorems from the arithmetic geometry literature, especially Breuil's classification, Chen/Gleason-Lim-Xu generic monodromy, Gleason-Lourenco connectedness, and Hodge-type integral models and uniformization. No new entities such as particles, forces, or dimensions are introduced. No numbers are fitted to data. The main internal gap is the propagation lemma Proposition 2.7, whose proof is sketched rather than completed.

assumptions (6)
  • domain assumption The local Shimura datum satisfies HN-irreducibility: the Newton polygon and Hodge polygon touch only at endpoints.
    This hypothesis is used in Theorem 3.5 to ensure generic points have large monodromy and in Theorem 1.5. It is stated explicitly but is a strong restriction.
  • domain assumption The Shimura datum is of Hodge type with G unramified at p, and the level at p is hyperspecial.
    This is needed for the canonical integral models of Kisin and Vasiu and for the Rapoport-Zink spaces of Hodge type.
  • standard math Breuil's classification of finite flat p-groups over O_K holds for p >= 3.
    Lemma 2.5 and Proposition 2.7 rely on Breuil's anti-equivalence of categories. This is a substantial external theorem, not proved in the paper.
  • standard math The theorem of Chen and of Gleason-Lim-Xu that generic points of the period domain have large monodromy.
    Quoted as Theorem 3.5, this is the source of the large-monodromy property for generic points. The paper does not reprove it.
  • standard math Gleason-Lourenco connectedness of p-adic period domains, used in Lemma 3.1.
    This external theorem is used to show all connected components of the Rapoport-Zink space have the same image under the period map.
  • standard math Kisin's integral models, Kim's Rapoport-Zink spaces and uniformization theorems for Hodge type, and Howard-Pappas constructions.
    These results are imported to identify formal neighborhoods and to pass between deformation spaces and Rapoport-Zink spaces.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards a p-adic nowhere density conjecture of Hecke Orbits." pith.science (2026). https://pith.science/paper/CQENLUN5

@misc{pith2026250606932,
  author       = {Pith},
  title        = {Pith review of: Towards a p-adic nowhere density conjecture of Hecke Orbits},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CQENLUN5}},
  note         = {Machine review of arXiv:2506.06932}
}
abstract

We propose a conjecture on the $p$-adic nowhere density of the Hecke orbit of subvarieties of Hodge type Shimura varieties. By investigating the monodromy of $p$-adic Galois representations associated with points on such Shimura varieties, we prove that the locus in a formal $\mathcal{O}_K$-neighborhood of a mod $p$ point that has large monodromy is open dense, where $K$ is a totally ramified finite extension of $\breve{\mathbb{Q}}_p.$

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Monodromy Groups of Supersingular Abelian Surfaces over $\mathbb{Q}_p$

    math.NT 2025-11 accept novelty 7.0 of 10

    For p≥7, the neutral components of p-adic monodromy groups of supersingular abelian surfaces over Q_p are classified and shown to be generically isomorphic to GL_2 ×_det GL_2.

Reference graph

Works this paper leans on

8 extracted references · 7 canonical work pages · cited by 1 Pith paper

  1. [1]

    [Bou89] N

    [Ber74] Pierre Berthelot, Cohomologie cristalline des schemas de caracteristique p>0, Lecture Notes in Maths 407 (1974). [Bou89] N. Bourbaki,Lie groups and lie algebras: Chapters 1-3, Bourbaki, Nicolas: Elements of mathematics, Springer,

    work page 1974

  2. [7]

    15 [Vas0308] Adrian Vasiu,Integral canonical models of shimura varieties of preabelian type, Asian J. Math. 3 (200308). [Vas0801] , Good reductions of shimura varieties of hodge type in arbitrary unramified mixed characteristic, part ii(200801). [Vas2010] , Good reductions of shimura varieties of hodge type in arbitrary unramified mixed characteristic, pa...

    work page 2021

  3. [764]

    [Duk8802] William Duke, Hyperbolic distribution problems and half-integral weight maass forms, Inventiones mathematicae92 (198802), 73–90

    [CU04] L.ClozelandE.Ullmo, équidistribution des points de hecke,ContributiontoAutomorphic Forms, Geometry, and Number Theory (2004). [Duk8802] William Duke, Hyperbolic distribution problems and half-integral weight maass forms, Inventiones mathematicae92 (198802), 73–90. [DvH22] Marco D’Addezio and Pol van Hoften,Hecke orbits on shimura varieties of hodge type,

    work page 2004

  4. [1989]

    2, 489–549

    [Bre00] Christophe Breuil,Groupes p-divisibles, groupes finis et modules filtres, Annals of Math- ematics 152 (2000), no. 2, 489–549. [CF00] Pierre Colmez and Jean-Marc Fontaine,Construction des représentations p-adiques semi- stables, Inventiones mathematicae140 (2000), no. 1, 1–43. [Cha05a] Ching-Li Chai, Hecke orbits on siegel modular varieties (Fedor ...

    work page 2000

  5. [2005]

    Chen, Composantes connexes géométriques de la tour des espaces de modules de groupes p-divisibles, Annales Scientifiques de l’Ecole Normale Superieure47 (2014), 723–

    [Che14] P.M. Chen, Composantes connexes géométriques de la tour des espaces de modules de groupes p-divisibles, Annales Scientifiques de l’Ecole Normale Superieure47 (2014), 723–

    work page 2014

  6. [2020]

    [Zha0501] Shou-Wu Zhang, Equidistribution of cm-points on quaternion shimura varieties , International Mathematics Research Notices 2005 (200501), no

    [Yu06] Chia-Fu Yu,Discrete hecke orbit problem for hilbert-blumenthal varieties, NCTS/TPE- Math Technical Report20 (2006), 1–20. [Zha0501] Shou-Wu Zhang, Equidistribution of cm-points on quaternion shimura varieties , International Mathematics Research Notices 2005 (200501), no. 59, 3657–3689, available at https://academic.oup.com/imrn/article-pdf/2005/59...

    work page 2006

  7. [2022]

    2, 621–645

    [EY03] Bas Edixhoven and Andrei Yafaev,Subvarieties of shimura varieties, Annals of Mathe- matics 157 (2003), no. 2, 621–645. [FM87] J.-M. Fontaine and W. Messing,P-adic periods and p-adic étale cohomology, Contempo- rary Math. 67 (1987), 179–207. [GL22] Ian Gleason and João Lourenço,On the connectedness ofp-adic period domains, arXiv e-prints (October 20...

    arXiv 2003

  8. [2023]

    3, 627–696, available atmath/0605254

    [Har13] Urs Hartl,On a conjecture of Rapoport and Zink, Inventiones Mathematicae193 (Sep- tember 2013), no. 3, 627–696, available atmath/0605254. [HP17] Benjamin Howard and Georgios Pappas,Rapoport–zink spaces for spinor groups, Com- positio Mathematica 153 (2017), no. 5, 1050–1118. [KIM18a] WANSU KIM, Rapoport–zink spaces of hodge type, Forum of Mathemat...

    work page arXiv 2017

Pith tools

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