Pith. sign in

REVIEW 3 major objections 6 minor 30 references

On the classification of indecomposable Ekedahl-Oort strata in unitary Shimura varieties, and related Newton polygons

T0 review · 3 major / 6 minor · reviewed 2026-08-02 · deepseek-v4-flash

Pith's one-line read Every indecomposable Ekedahl-Oort stratum of GU(a,b) over an odd inert prime is one of four word-built types, and tautological lifts place each stratum on its Newton stratum.

desk verdict Solid four-type classification, but a real signature-convention bug affects the examples and the word-to-Weyl algorithm — worth reviewing after a fix. read the letter →

arxiv 2606.16882 v2 pith:4OX6WREU submitted 2026-06-15 math.NT

classification math.NT MSC 11G1814G3511G1014L15
keywords Ekedahl-OortstratificationunitaryShimuravarietiesBT1modulesDieudonnéKraftdiagramsNewtonpolygonsWeylgroupcosetssupersingularlocus
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

Ekedahl-Oort strata partition the characteristic-p fiber of a Shimura variety by the isomorphism type of the p-torsion of the abelian varieties it parameterizes. This paper aims to classify the indecomposable Ekedahl-Oort strata of unitary Shimura varieties GU(a,b) over an odd prime that is inert in the quadratic field. The claimed classification is that every unitary BT1 module—the linear-algebraic avatar of unitary p-torsion—decomposes uniquely into four building blocks: unitary unicycles, unitary bicycles, Serre unicycles, and Serre bicycles, so the indecomposable strata are exactly these four types. The paper also gives an algorithm translating the word description of a stratum into a Weyl-group coset representative, and constructs a tautological lift of each stratum whose Newton polygon is read off from the words. From this it follows that every stratum constructed this way intersects the corresponding Newton stratum, and the unicycle-type strata always meet the supersingular locus.

What carries the argument

The load-bearing objects are unitary BT1 modules: polarized mod-p Dieudonné modules (finite-dimensional k-vector spaces with Frobenius F and Verschiebung V satisfying FV=VF=0 and ker F = im V) equipped with a symplectic pairing and a decomposition M0⊕M1 such that F and V swap the summands and the summands are Lagrangian. The organizing tool is the word construction: every indecomposable polarized BT1 module is M(w), built from a primitive word w in {f,v} by drawing a Kraft diagram; the unitary condition forces the four special word shapes above. The Serre tensor construction STC(N)=N0⊕N1 turns an arbitrary polarized BT1 module into a unitary one and is exactly what produces the two 'Serre' t

What would settle it

Find an indecomposable unitary BT1 module of signature (1,1) other than the Serre bicycle described in the paper; since the paper asserts (1,1) has exactly one indecomposable type, such an object would overturn the classification.

Watch

Extended reading notes

Core claim

The paper's central theorem (4.21) states that any unitary BT1 module of signature (a,b), equivalently any Ekedahl-Oort stratum of GU(a,b), decomposes uniquely as a direct sum of unitary unicycles, unitary bicycles, Serre unicycles, and Serre bicycles. The only indecomposable types are therefore these four: primitive words of the form yy* with y odd (unitary unicycle); primitive non-self-dual even-length words (unitary bicycle); primitive self-dual words whose length is divisible by 4 (Serre unicycle); and words zz with z primitive of odd length (Serre bicycle). The proof reduces the unitary structure to word combinatorics via a standard-object theorem giving a basis B0∪B1 preserved by F and

Load-bearing premise

The classification depends on a cited standard-object theorem: every unitary BT1 module has a basis B0∪B1 preserved by F and V (making all loop words even) and a unitary pairing unique up to isomorphism; if either fails, the four-type decomposition does not follow.

Editorial extensions

If this is right

  • The Ekedahl-Oort stratification of GU(a,b) is completely understood at the level of indecomposable strata in every signature: the four word types exhaust all possibilities.
  • The word-to-Weyl algorithm makes the geometry of each stratum (dimension, closure relations through the Bruhat order) accessible from the word combinatorics.
  • For every multiset of words W satisfying the signature conditions, the Ekedahl-Oort stratum indexed by γ(W) meets the Newton stratum cut out by NP(W).
  • Unitary unicycle and Serre unicycle strata always contain supersingular points.
  • In odd total dimension q the only indecomposable strata are unitary unicycles; in even q the remaining three types appear, with Serre types confined to parallel signature.

Reading between the lines

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

  • Beyond the paper: the balanced-word criterion in the supersingularity corollary is sufficient but not necessary for a stratum to meet the supersingular locus, since the paper's own later example gives a non-balanced word whose stratum still contains a supersingular point; a complete combinatorial characterization would need extra bookkeeping beyond word slopes.
  • An implicit consequence is that counting indecomposable strata of a given signature reduces to counting rotation orbits of primitive words satisfying the four shape conditions, which could be made explicit for small a,b and compared with known coset counts.
  • The same word-slope tautological-lift construction may apply to other PEL Shimura varieties whose BT1 modules admit an analogous word decomposition, potentially yielding EO-Newton intersection results outside Coxeter-type settings.
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 / 6 minor

Summary. The paper gives a word-based classification of indecomposable unitary BT1 modules in arbitrary signature (a,b) over an algebraically closed field of characteristic p>2. The main theorem (Thm. 4.21) asserts that every unitary BT1 module decomposes uniquely as a direct sum of unitary unicycles, unitary bicycles, Serre unicycles, and Serre bicycles. Section 5 provides an algorithm converting the word description of a stratum into a minimal Weyl-group coset representative (Thm. 5.3). Section 6 constructs a tautological lift of each unitary BT1 module, computes its Newton polygon (Prop. 6.6), and proves that the corresponding Ekedahl-Oort stratum intersects the corresponding Newton stratum (Thm. 6.10); corollaries identify strata that meet the supersingular locus. The arguments use the Kraft word construction, Oort's polarized BT1 structure theorem, Moonen's standard-object theorem, and Viehmann-Wedhorn's realization theorem.

Significance. If the central classification is established, this is a substantial and useful contribution: it gives a complete, explicit, combinatorial description of indecomposable Ekedahl-Oort strata for unitary Shimura varieties in arbitrary signature, together with a transparent word-to-Weyl dictionary and Newton-polygon computations. The paper is commendably explicit: it supplies concrete pairings, detailed examples, and a constructive tautological lift, and it clearly flags its reliance on Moonen's and Viehmann-Wedhorn's theorems. The main results are falsifiable and well illustrated. The proofs of two load-bearing lemmas are, however, too compressed for the central claims to be considered fully established as written.

major comments (3)
  1. [§4.4, Lemma 4.18] Lemma 4.18 is load-bearing for Equation (4.B) and hence for the structure theorem 4.21, but its proof is only two sentences. It must justify: (i) the existence of a basis element e_{i,j} in M(w) that is λ-orthogonal to all of M(w) when λ|M(w) is degenerate; (ii) why any nonzero pairing partner of this element must lie in a copy of M(w*), rather than merely in some other summand; and (iii) why the multiplicity of M(w*) equals that of M(w), not just that at least one partner exists. The present sketch does not address multiplicity matching. This needs a full argument or a precise reference to a general duality statement.
  2. [§4.2, Lemma 4.4] The 'only if' direction of Lemma 4.4 concludes that q=ℓ(y) is odd because 'e_{0,1} can only pair non-trivially with the element … q steps away.' This is not proved. Lemma 3.7 constructs one pairing on a self-dual word, but uniqueness of the symplectic pairing on the indecomposable M(w) up to isomorphism is not established before Lemma 4.4. A rigorous proof should show that any λ satisfying λ(Fx,y)=λ(x,Vy)^p is equivalent to the pairing of Lemma 3.7. This point is essential because the oddness of ℓ(y) is what makes the even/odd decomposition M0/M1 Lagrangian and hence produces unitary unicycles.
  3. [§5.2, Theorem 5.3] The word-to-Weyl algorithm is central to identifying the Ekedahl-Oort stratum in Theorem 6.10, but the proof is compressed in three places. (a) It is not shown explicitly that q_W equals the rank of M0 for the module under consideration. (b) The lexicographic ordering of the multiset eW is used as if canonical, but equal words from repeated τ^2-orbits or multiplicities require a consistent tie-breaking rule; the proof should show that γ(W) is independent of this tie-breaking and of the non-unique decomposition in Equation (4.A). (c) The step identifying the canonical filtration of [PU21, §3.5] with the lexicographic order of the words w_{i,j} raised to the power L_W/ℓ_i needs a more detailed verification. These details are necessary for the dictionary to be reliable.
minor comments (6)
  1. [§3.1, Rem. 4.7, Ex. 4.8, Ex. 4.16] The apparent swap of a and b in the examples disappears if one reads u_j as in §3.1, where u_0 is the rightmost letter. Please add an explicit sentence to Remark 4.7 and Remark 4.15 stating this indexing convention; without it, the formulas can easily be misread in the opposite direction.
  2. [§4.3, Lemma 4.9] The phrase 'where q is even' is confusing because q was previously used for a+b in Definition 4.1 and for ℓ(y) in Lemma 4.4; in Lemma 4.9 q appears to be the length of w. Please use a different letter for the length of a word to avoid ambiguity.
  3. [§4.5, Proposition 4.28] Proposition 4.28 is dismissed with 'This can be verified with the proof method of Proposition 4.27.' Since this proposition underpins the Serre-bicycle terminology and is used in Remark 6.14, please include at least a summary of the isomorphism or a precise reference.
  4. [§6.2, Proposition 6.4] The chain of equalities in the signature computation contains a typo: 'dim_k(M0/F(M1)) = dim_k(M0/F(M1))' should read 'dim_k(M0/F(M1)) = q - dim_k(F(M1)) = dim_k(M1[F]) = a.'
  5. [§6.4, Theorem 6.10] The only passage from a constructed p-divisible group to a point of M(a,b) is [VW13, Thm 1.6(2)]. This dependence should be stated explicitly in the statement or proof of Theorem 6.10, since it is a nontrivial realization theorem.
  6. [Rem. 3.2, reference [Moo96]] Reference [Moo96] is identified as E. H. Moore, 'A two-fold generalization of Fermat's theorem.' The linear-independence fact over F_p cited to page 196 may not be the intended source; please verify and correct the reference.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity; the classification and Newton-intersection results are derived from Moonen, Kraft/Oort, and VW13, not from the paper's own conclusions.

full rationale

The derivation chain is not circular. Theorem 4.21 builds on Kraft's word decomposition (Prop. 3.3), Oort's polarized decomposition (Prop. 3.11), Moonen's standard-object basis and uniqueness of the pairing ([Moo01, (4.9), (6.7)]), with Lemmas 4.17-4.20 supplying the unitary refinements. None of these inputs is an earlier statement of the paper's four-type theorem. The word-to-Weyl algorithm (Definition 5.1, Theorem 5.3) is proved from the canonical filtration of [PU21, Section 3.5] and Moonen's correspondence; its z_j's are computed from the words, not chosen to match a preassigned coset. Proposition 6.6 derives NP(W) from F^ell = p^{#v} on each tautological lift, and Theorem 6.10 uses [VW13, Thm 1.6(2)] to realize the lift geometrically; no parameter is fitted to the desired Newton polygon. The self-citations [ABF+24], [ABF+25], and [GLP26] are for notation, background, and a counterexample, and none is load-bearing. There is, however, a non-circular correctness concern: Definition 4.1 (and Definition 6.2) define a = dim M1[F], while Remark 4.7 and Example 4.16 compute signatures with the roles of M0[F] and M1[F] interchanged (e.g., ffvffvvfvv is said to have signature (3,2), whereas the paper's own Definition 4.1 gives (2,3)). This affects the a/b labeling of the claimed W(a,b) bijection, but it is a convention/calculation issue, not a reduction of the results to their inputs.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

No numerical parameters are fitted to data. The only auxiliary choices are arbitrary units c and C used to normalize polarizations (c^p=-c, C^σ=-C); these do not affect isomorphism classes. The central claim rests on published structural theorems: Kraft's decomposition, Oort's polarized classification, Moonen's standard-object/uniqueness theorem, Pries-Ulmer's canonical filtration, Viehmann-Wedhorn's realization theorem, and Serre tensor construction properties. No new particles, forces, or physical entities are introduced; 'unicycle' and 'bicycle' are terminology for existing mathematical objects. The main dependence not proved in this paper is Moonen's basis/pairing theorem, which is the load-bearing external input.

assumptions (7)
  • standard math Kraft structure theorem: every BT1 module is a finite direct sum of cyclic modules M(w) attached to primitive words in {f,v}.
    Cited from unpublished [Kra75] via Prop 3.3; base of all word decompositions in Sections 3-5.
  • standard math Oort's classification of polarized BT1 modules: indecomposables are unicycles and bicycles (Prop 3.11, [Oor01, 9.9]).
    Used in Section 3 and to frame the unitary analogue in Theorem 4.21.
  • standard math Moonen's standard-object theorem: every unitary BT1 module has a basis B0 ∪ B1 preserved by F and V, and its pairing is unique up to isomorphism ([Moo01, (4.9), (6.7)]).
    Appears before Eq. (4.A), Lemma 4.20, Prop 4.27, and Theorem 5.3; if false, the word decomposition and uniqueness proofs collapse.
  • standard math Pries-Ulmer canonical filtration [PU21, Sec. 3.5] computes the final filtration W• used to extract Weyl coset representatives.
    Needed for Theorem 5.3; cited but not reproduced.
  • domain assumption Viehmann-Wedhorn [VW13, Thm 1.6(2)]: every unitary p-divisible group over k with the given signature occurs as A[p^∞] for a point of M(a,b) with level structure.
    Bridges the module-theoretic tautological lift in Theorem 6.10 to an actual point of the Shimura variety.
  • domain assumption Serre tensor construction: for a polarized abelian variety A, O_K ⊗_Z A exists with OK-action, polarization h⊗λ, and p-torsion F_{p^2} ⊗_{F_p} A[p] ([AK18, Con04, CCO14, Lau02]).
    Provides the geometric meaning of Serre unicycles and bicycles and is used in Remark 6.14.
  • domain assumption Dieudonné theory with contravariant convention as set up in Section 2.3, with p>2 inert in K and k algebraically closed.
    All module/group-scheme equivalences in the paper use this convention.

how reviews work

0 comments
Cite this review

Pith. "Pith review of On the classification of indecomposable Ekedahl-Oort strata in unitary Shimura varieties, and related Newton polygons." pith.science (2026). https://pith.science/paper/4OX6WREU

@misc{pith2026260616882,
  author       = {Pith},
  title        = {Pith review of: On the classification of indecomposable Ekedahl-Oort strata in unitary Shimura varieties, and related Newton polygons},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4OX6WREU}},
  note         = {Machine review of arXiv:2606.16882}
}
abstract

In this paper, we give a complete classification of indecomposable Ekedahl--Oort strata of Shimura varieties associated to the unitary group $\mathsf{GU}(a, b)$ over an odd inert prime. We show that each indecomposable stratum is one of four types: unitary unicycle, unitary bicycle, Serre unicycle, or Serre bicycle; the latter two types are named for a tensor construction of abelian varieties developed by Serre. We provide an algorithm that translates the description of a stratum in terms of words in the alphabet $\{\texttt{f},\texttt{v}\}$ to the corresponding Weyl group coset representative. Finally, using a $p$-adic lift, we construct a `tautological' point in each Ekedahl--Oort stratum, and compute its Newton polygon. As an application, we show that the indecomposable Ekedahl--Oort strata corresponding to unitary unicycles and Serre unicycles always intersect the supersingular locus.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

30 extracted references · 2 linked inside Pith

  1. [1]

    Ekedahl- O ort strata in the GU (q-2,2) S himura variety

    Emerald Andrews, Deewang Bhamidipati, Maria Fox, Heidi Goodson, Steven Groen, and Sandra Nair. Ekedahl- O ort strata in the GU (q-2,2) S himura variety. arXiv e-prints , 2024. arxiv:2405.04464

  2. [2]

    The E kedahl- O ort and N ewton stratification of the GU (3,2) S himura variety

    Emerald Andrews, Deewang Bhamidipati, Maria Fox, Heidi Goodson, Steven Groen, and Sandra Nair. The E kedahl- O ort and N ewton stratification of the GU (3,2) S himura variety. arXiv e-prints , 2025

  3. [3]

    Serre's tensor construction and moduli of abelian schemes

    Zavosh Amir-Khosravi. Serre's tensor construction and moduli of abelian schemes. Manuscripta Math. , 156(3-4):409--456, 2018

  4. [4]

    Achter and Rachel Pries

    Jeffrey D. Achter and Rachel Pries. Superspecial rank of supersingular abelian varieties and J acobians. J. Th\'eor. Nombres Bordeaux , 27(3):605--624, 2015

  5. [5]

    Combinatorics of C oxeter groups

    Anders Bj\"orner and Francesco Brenti. Combinatorics of C oxeter groups . Graduate Texts in Mathematics. Springer, New York, 2005

  6. [6]

    Congruence relations for S himura varieties associated to some unitary groups

    Oliver B\"ultel and Torsten Wedhorn. Congruence relations for S himura varieties associated to some unitary groups. J. Inst. Math. Jussieu , 5(2):229--261, 2006

  7. [7]

    Complex multiplication and lifting problems , volume 195 of Mathematical Surveys and Monographs

    Ching-Li Chai, Brian Conrad, and Frans Oort. Complex multiplication and lifting problems , volume 195 of Mathematical Surveys and Monographs . American Mathematical Society, Providence, RI, 2014

  8. [8]

    Gross- Z agier revisited

    Brian Conrad. Gross- Z agier revisited. In Heegner points and R ankin L -series , volume 49 of Math. Sci. Res. Inst. Publ. , pages 67--163. Cambridge Univ. Press, Cambridge, 2004. With an appendix by W. R. Mann

Show all 30 references
  1. [9]

    Lectures on p -divisible groups

    Michel Demazure. Lectures on p -divisible groups . Lecture Notes in Mathematics. Springer-Verlag, Berlin, 1986. Reprint of the 1972 original

  2. [10]

    Basic loci of C oxeter type in S himura varieties

    Ulrich G\" o rtz and Xuhua He. Basic loci of C oxeter type in S himura varieties. Camb. J. Math. , 3(3):323--353, 2015

  3. [11]

    Basic loci of C oxeter type with arbitrary parahoric level

    Ulrich G\"ortz, Xuhua He, and Sian Nie. Basic loci of C oxeter type with arbitrary parahoric level. Canad. J. Math. , 76(1):126--172, 2024

  4. [12]

    Groen, Elvira Lupoian, and Mychelle Parker

    Steven R. Groen, Elvira Lupoian, and Mychelle Parker. Intersections of the E kedahl- O ort and N ewton strata of A _ 5 . arXiv e-prints , 2026. arxiv:2509.19878

  5. [13]

    Generic N ewton polygons of E kedahl- O ort strata: O ort's conjecture

    Shushi Harashita. Generic N ewton polygons of E kedahl- O ort strata: O ort's conjecture. Ann. Inst. Fourier , 60(5):1787--1830, 2010

  6. [14]

    The G -stable pieces of the wonderful compactification

    Xuhua He. The G -stable pieces of the wonderful compactification. Trans. Amer. Math. Soc. , 359(7):3005--3024, 2007

  7. [15]

    On the supersingular locus of the GU (2,2) S himura variety

    Benjamin Howard and Georgios Pappas. On the supersingular locus of the GU (2,2) S himura variety. Algebra Number Theory , 8(7):1659--1699, 2014

  8. [16]

    Kottwitz

    Robert E. Kottwitz. Points on some S himura varieties over finite fields. J. Amer. Math. Soc. , 5(2):373--444, 1992

  9. [17]

    Kommutative algebraische p - G ruppen (mit A nwendungen auf p -divisible G ruppen und abelsche V ariet\" a ten)

    Hanspeter Kraft. Kommutative algebraische p - G ruppen (mit A nwendungen auf p -divisible G ruppen und abelsche V ariet\" a ten). Manuscript, September 1975

  10. [18]

    The maximum or minimum number of rational points on genus three curves over finite fields

    Kristin Lauter. The maximum or minimum number of rational points on genus three curves over finite fields. Compositio Math. , 134(1):87--111, 2002. With an appendix by Jean-Pierre Serre

  11. [19]

    A two-fold generalization of F ermat's theorem

    Eliakim Hastings Moore. A two-fold generalization of F ermat's theorem. Bull. Amer. Math. Soc. , 2(7):189--199, 1896

  12. [20]

    Group schemes with additional structures and W eyl group cosets

    Ben Moonen. Group schemes with additional structures and W eyl group cosets. In Moduli of A belian V arieties , volume 195 of Progr. Math. , pages 255--298. Birkh\"auser, Basel, 2001

  13. [21]

    A stratification of a moduli space of abelian varieties

    Frans Oort. A stratification of a moduli space of abelian varieties. In Moduli of A belian V arieties , volume 195 of Progr. Math. , pages 345--416. Birkh\"auser, Basel, 2001

  14. [22]

    Simple p -kernels of p -divisible groups

    Frans Oort. Simple p -kernels of p -divisible groups. Adv. Math. , 198(1):275--310, 2005

  15. [23]

    On BT_1 group schemes and F ermat curves

    Rachel Pries and Douglas Ulmer. On BT_1 group schemes and F ermat curves. New York J. Math. , 27:705--739, 2021

  16. [24]

    Algebraic zip data

    Richard Pink, Torsten Wedhorn, and Paul Ziegler. Algebraic zip data. Doc. Math. , 16:253--300, 2011

  17. [25]

    On the supersingular locus of the GU(2,n-2) S himura variety

    Ryosuke Shimada. On the supersingular locus of the GU(2,n-2) S himura variety. arXiv , page 2410.05110, 2024

  18. [26]

    The supersingular locus of the S himura variety for GU (1,s)

    Inken Vollaard. The supersingular locus of the S himura variety for GU (1,s) . Canad. J. Math. , 62(3):668--720, 2010

  19. [27]

    The supersingular locus of the S himura variety of GU (1,n-1) II

    Inken Vollaard and Torsten Wedhorn. The supersingular locus of the S himura variety of GU (1,n-1) II . Invent. Math. , 184(3):591--627, 2011

  20. [28]

    Ekedahl- O ort and N ewton strata for S himura varieties of PEL type

    Eva Viehmann and Torsten Wedhorn. Ekedahl- O ort and N ewton strata for S himura varieties of PEL type. Math. Ann. , 356(4):1493--1550, 2013

  21. [29]

    Specialization of F -zips

    Torsten Wedhorn. Specialization of F -zips. arXiv e-prints , 2005. arxiv:0507175

  22. [30]

    The Ekedahl-Oort stratification of unitary Shimura varieties

    Amy Wai Ling Jane Wooding. The Ekedahl-Oort stratification of unitary Shimura varieties . PhD thesis, McGill University, 2016

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.