Unveiling Crystal Embeddings: New Perspectives on String Polytopes and Atomic Decompositions

Jacinta Torres; Lara Bossinger

REVIEW 4 major objections 5 minor 22 references

Unveiling Crystal Embeddings: New Perspectives on String Polytopes and Atomic Decompositions

T0 review · 4 major / 5 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 shows that in type A_{n-1}, translating a string polytope by the extreme weight-zero vectors z_1 and z_{n-1} is an injective crystal morphism S_i(λ) → S_i(λ+θ), and conjectures these translations give n−1 distinct atomic…

desk verdict Clean idea and a useful conjecture, but the proof of the main crystal-compatibility theorem is a sketch with an unproved path-counting assertion and a sign issue that needs real referee attention. read the letter →

arxiv 2505.22127 v1 pith:LFO7NII5 submitted 2025-05-28 math.RT math.CO

Lara Bossinger , Jacinta Torres This is my paper

classification math.RTmath.CO MSC 17B3705E10

keywords stringpolytopescrystalbasesatomicdecompositionswiringdiagramsGleizer–PostnikovpathshighestroottypeAembeddings

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

String polytopes are polyhedral models whose lattice points parameterise the strings of type-A crystal bases. This paper constructs n−1 embeddings of the string polytope S_i(λ) into S_i(λ+θ), where θ is the highest root, by translating every lattice point by one of n−1 distinguished vectors z_j of weight zero. For the two extreme vectors z_1 and z_{n-1}, the translation commutes with the crystal operators wherever they are defined, so it is an injective crystal morphism. The intermediate embeddings become compatible with the crystal structure after iterated projections that remove the first or last wire from the wiring diagram. Building on these embeddings, the paper conjectures a way to obtain n−1 different atomic decompositions of the crystal B(kθ), a problem for which only one decomposition was previously known.

What carries the argument

The load-bearing machinery is the wiring-diagram (pseudoline) model of string cones: for a fixed reduced expression i, the string cone S_i is cut out by inequalities indexed by Gleizer–Postnikov rigorous paths, one for each orientation of a pair of adjacent wires. The same paths, through the integer vectors r(γ) and s(γ), give the formulas for ε_a, f_a, and e_a on lattice points of S_i(λ). The distinguished vectors z_1,...,z_{n-1} are the weight-zero vertices of the adjoint crystal B(θ); their string coordinates mark the first crossings along boundary wires, so the pairing ⟨r(γ), z_j⟩ is 1, −1, or 0 depending on how γ meets the boundary. This path-counting is what makes the translation x ↦ x+z_j shift ε and φ values in the controlled way that defines a crystal morphism.

What would settle it

Enumerate all left Gleizer–Postnikov paths in Γ_1 for the reduced expression i=(1,2,3,4,1,2,3,1,2,1) in type A_4 and count turning points on wire ℓ_1; finding a path with zero or two such points, and then checking that ε_1(x+z_1) differs from ε_1(x)+1 for the corresponding lattice point x, would refute Theorem 4.3(1).

Watch

Extended reading notes

Core claim

The central discovery is that the weight-zero elements z_1,...,z_{n-1} of the adjoint crystal B(θ) act on every string polytope S_i(λ) by translation, and that the two boundary elements z_1 and z_{n-1} are crystal morphisms. On the wiring-diagram model, z_1 is supported on the crossings of the bottom wire ℓ_1 and z_{n-1} on the crossings of the top wire ℓ_n; Theorem 4.3 shows that translating by z_1 or z_{n-1} raises ε_1 (respectively ε_{n-1}) by exactly 1 whenever it is positive, leaves all other ε_a and φ_a unchanged, and preserves the analogous φ-values, so the translation x ↦ x+z_j sends S_i(λ) into S_i(λ+θ) as an injective crystal morphism. Corollary 4.6 extends a weaker compatibility to all j via iterated projections onto smaller-rank string cones, and Example 5.5 shows the proposed atom agrees with the known big atom in type A_2.

Load-bearing premise

The proof of the crystal-morphism statement for the extreme embeddings assumes, without proof, that every rigorous path in the boundary families Γ_1 and Γ_{n−1} has exactly one turning point on the boundary wire it touches; if some path had zero or two such turning points, the equality ε_a(x+z_1)=ε_a(x)+1 would fail and the morphism property would break.

Editorial extensions

If this is right

For j=1 and j=n−1, the translation x ↦ x+z_j gives an embedding S_i(λ) ↪ S_i(λ+θ) for every dominant λ and every reduced expression i of type A_{n−1}.
Because the translation preserves the positivity of ε_a and φ_a, every crystal operator defined on a lattice point of the domain has a well-defined image under the embedding, so the embedding is a genuine crystal morphism.
For the intermediate j, compatibility with crystal operators is recovered after repeatedly projecting away the first or last wire, reducing to the extreme case in a smaller rank.
The proposed atoms are complements of unions of the images of the n−1 embeddings inside B(kθ); if Conjecture 5.3 holds, each j gives one atomic decomposition, yielding n−1 of them.
In type A_2, the constructed atom coincides with the previously known big atom, so the conjecture is compatible with the one known atomic decomposition.

Reading between the lines

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

The proof of Theorem 4.3 rests on an unproved path-counting assertion about turning points on boundary wires; a general proof of that assertion would let the same translation-by-z_j mechanism be exported to other classical types, where atomic decompositions are largely open.
The companion observation that the extreme embedding adds a fixed charge of 1 to the classical charge statistic of a tableau suggests that these embeddings could yield a direct polyhedral proof of positive formulas for q-analogues of weight multiplicities in type A.
A computer search over all rigorous paths for small n could verify the boundary path-counting lemma and test Conjecture 5.3 for n ≤ 5 and several multiples k; either outcome would clarify how much of the atomic-decomposition pattern is combinatorial rather than representation-theoretic.
The same construction might be iterated: after embedding S_i(λ) into S_i(λ+θ), translating again by the same z_j embeds into S_i(λ+2θ), producing a graded chain of crystals whose limit could be a model for the principal nilpotent or pre-canonical bases.

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

Short version: the paper's genuinely new content is Theorem 4.3 (crystal compatibility for the two extremal translation embeddings) and Conjecture 5.3 (a family of atomic decompositions of B(kθ) in type A). The embeddings themselves are just translations by zero-weight points, which is simple; the interest is whether they are crystal morphisms. The sl(3) example that matches the known Patimo/Lenart-Lecouvey atom is good evidence, and the paper is honest about what is conjectural.

The soft spot is the proof of Theorem 4.3. In the a=1=i case, the proof asserts that every rigorous path in Γ_1 has exactly one turning point on the boundary wire ℓ_1. That is not proved. The rigorous-path definition does not obviously exclude a path that leaves ℓ_1 and later re-enters it. If such a path exists, the claimed increment ε_1(x+z_1)=ε_1(x)+1 becomes +2, and the embedding would not land in S_i(λ+θ). The a=n−1 case has the same problem. The a≠i case also reads wrong: the proof says the two turning points on ℓ_i contribute opposite signs, but under the paper's own convention (Lemma 4.1, z_1 supported on v_{1,j}, j>1) both have sign +1, so there is no cancellation. This might be fixable with a consistent ordered-pair convention, but as written it does not work.

Corollary 4.6 is likewise only a one-sentence appeal to the crystal graph of V(θ), without showing that the projections preserve the needed crystal data. Given that Theorem 4.3 is the load-bearing step, these gaps mean the central claim is not yet established.

The paper is still useful. The conjecture is concrete and motivated by computation, and the surrounding discussion of atomic decompositions is clear. Specialists in string polytopes and crystals for Kostka-Foulkes formulas will want to see this, and the idea deserves a serious referee. But the referee should insist on a complete proof of Theorem 4.3 and a proper argument for Corollary 4.6.

Referee Report

4 major / 5 minor

Summary. The paper introduces n-1 embeddings of string polytopes of type A, given by translations x ↦ x+z_j by zero-weight vectors of the adjoint crystal B(θ). For j=1 and j=n-1, the authors claim (Theorem 4.3, Corollary 4.5) that these translations induce injective crystal morphisms S_i(λ) → S_i(λ+θ) for any dominant weight λ and any reduced expression i of the longest Weyl group element. A further corollary (Corollary 4.6) asserts a weaker compatibility for all j via projections that remove boundary wires, and the paper concludes with a conjecture (Conjecture 5.3) on obtaining n-1 distinct atomic decompositions of B(kθ). The proof of Theorem 4.3 is based on the Gleizer-Postnikov rigorous-path description of the string cone and the Genz-Koshevoy-Schumann crystal formula.

Significance. If the main claims are correct, the paper provides new crystal embeddings for string polytopes in type A and proposes a concrete route toward new atomic decompositions, complementing the known constructions of Patimo and Lenart-Lecouvey. The paper correctly builds on prior work (Littelmann, Berenstein-Zelevinsky, Gleizer-Postnikov, Genz-Koshevoy-Schumann) and includes a testable conjecture supported by SageMath experiments and an explicit sl_3 example. However, the current manuscript does not rigorously establish its central theorem: the path-counting assertions in the proof of Theorem 4.3 are not proved, the statement of Theorem 4.3 is weaker than what Corollary 4.5 needs, and Corollary 4.6 is only sketched. These gaps are load-bearing for the paper's main claims, though they appear curable in a revision.

major comments (4)

[Section 4, Theorem 4.3 proof] The case a=i asserts without proof that every path γ∈Γ_1 has exactly one turning point on ℓ_1 (and similarly Γ_{n−1} on ℓ_n). This assertion is true: a path in Γ_1 must turn at the unique crossing of ℓ_1 and ℓ_2 and cannot leave ℓ_1 afterwards, since each pair of wires crosses only once. However, the proof does not provide this justification, and the argument is therefore incomplete. The case a≠i is also mishandled: the proof claims that a path containing a segment of ℓ_i has two turning points of opposite signs on that wire. Under the coordinate convention of Lemma 4.1(iii), where coordinates are indexed by i<j, every turning point on ℓ_1 would contribute +1, not opposite signs. In fact, no rigorous path in Γ_a for a≠1 can contain a segment of ℓ_1 at all; the proof should state this directly instead of the incorrect sign argument.
[Section 4, Corollary 4.5] Corollary 4.5 does not follow from the stated Theorem 4.3. The theorem only preserves positivity of ε_a and φ_a, but the corollary needs the stronger identities ε_a(x+z_i)=ε_a(x)+δ_{a,i} and φ_a(x+z_i)=φ_a(x)+δ_{a,i} (or at least the constant shift of the linear forms ⟨x,r(γ)⟩) to prove both that x+z_i lies in S_i(λ+θ) and that the translation is a crystal morphism. These equalities are claimed in the proof of Theorem 4.3 but are not stated in the theorem, and the corollary does not explain how they imply the morphism property. The authors should restate Theorem 4.3 with the exact identities and prove that the shift is constant across all γ∈Γ_a, so that the maximal (or minimal) paths are preserved and the operators e_a and f_a commute with the translation.
[Section 4, Corollary 4.6] This corollary is not proved. The projections pr_1 and pr_n are defined by removing the first or nth wire, but the manuscript does not establish that, for an arbitrary reduced expression i, the image pr(S_i(λ)) is the string polytope for sl_{n−1} associated with the projected word, nor that these projections are compatible with the crystal operators on the polytopes. The proof's appeal to the crystal graph of V(θ) only shows how the elements z_i project; it does not address the required compatibility of the projections with the polyhedral and crystal structure. This gap must be filled, or the statement should be explicitly reformulated as a conjecture.
[Section 5, Example 5.5] The map ψ^3_2 is defined as x↦(x_1,x_2+1,x_3+1) with i=(2,1,2). According to Lemma 4.1(ii), this is translation by z_1 (the greedy subword '1 2' gives (0,1,1)), not by z_2 (which is (1,1,0)). Since Corollary 4.5 defines ψ^n_i as translation by z_i, the subscript in ψ^3_2 is inconsistent. Please correct the label or, if a different convention is used, clarify it explicitly.

minor comments (5)

[Throughout] There are numerous typos: 'psudoline' (p.4), 'muust' (p.8), 'Rigurous' (p.5), 'instrinsic' (p.5), and 'Luztig' (p.1), among others. A careful proofreading pass is needed.
[Section 4] The symbol i is used both for a fixed reduced expression and for the index of z_i, which makes statements such as 'i∈{1,n−1}' in Theorem 4.3 ambiguous. Consider denoting the reduced expression by, for example, w or a different letter.
[Lemma 4.1] In the proof of Lemma 4.1, the distinctness of the z_j is asserted rather than demonstrated; the argument that the greedy subwords give distinct vertices should be expanded.
[Section 5, Conjecture 5.3] The text says 'ndistinct atomic decompositions' but the construction described in the conjecture yields n−1 decompositions; please correct this wording.
[References] Reference [20] appears to have an incorrect title; please check the bibliographic data against the published version.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the crystal-embedding theorem is proved from external path combinatorics and explicit coordinates; the unproved turning-point count is a rigor gap, not a self-referential step.

full rationale

The derivation chain is self-contained. Theorem 4.3 computes ε_a(x + z_i) and φ_a(x + z_i) from the external Genz–Koshevoy–Schumann formula ε_a(x) = max{⟨x, r(γ)⟩ : γ ∈ Γ_a} together with the explicit coordinate description of z_1 and z_{n−1} in Lemma 4.1. Lemma 4.1 is proved directly from the reduced word and the wiring diagram, not from the crystal-morphism statement being proved. The claimed embeddings are therefore not defined in terms of the target crystal morphism, and no fitted parameter is renamed as a prediction. Example 5.5 checks the resulting atom against the independently defined atoms of Patimo and Lenart–Lecouvey, which is an external benchmark rather than a restatement of the paper's own construction. The only self-citations, [2] and [22], appear in motivational remarks and do not supply any load-bearing step of the proof; in particular, no uniqueness theorem or ansatz is imported from the authors' prior work. The unproved assertion in Theorem 4.3 about paths in Γ_1 and Γ_{n−1} having exactly one turning point on the relevant boundary wire is a genuine combinatorial gap for correctness review, but it is not a circular definition, a fitted-input prediction, or a self-citation chain. The central claim has independent mathematical content, so the circularity score is 0.

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

No numerical parameters are fitted. The central claim rests on standard crystal and string-polytope theory plus two paper-specific combinatorial assumptions: the zero-weight coordinate description of z_i, and the behavior of the wire-removal projections. The latter is the most fragile.

assumptions (6)

standard math Standard theory of crystals and tensor products, cited to Bump-Schilling [3].
Used throughout Section 2 to define crystals, tensor products, and the highest weight crystal B(λ).
standard math Littelmann and Berenstein-Zelevinsky description of string cones and polytopes.
Defines S_i(λ) and identifies lattice points with adapted strings; cited as Theorem 2.4 and [18, 1].
standard math Gleizer-Postnikov rigorous path description of string cones.
Used in Section 3 to encode the inequalities of S_i and S_i(λ); cited as Theorem 3.4 and [10].
standard math Genz-Koshevoy-Schumann explicit crystal operators via rigorous paths.
Used in the proof of Theorem 4.3 to compute ε_a and φ_a; cited as Theorem 3.7 and [9].
domain assumption Lemma 4.1: the weight-zero elements of B_i(θ) are exactly z_1,...,z_{n−1} with explicit greedy subword coordinates.
Load-bearing for the translation maps and for the sign computations in Theorem 4.3; the proof is terse and partly combinatorial.
ad hoc to paper Projections pr_j, obtained by removing the first or nth wire, map string polytopes for sl_n to string polytopes for sl_{n−1} and are compatible with crystal operators.
Corollary 4.6 assumes this without proof; the paper only illustrates pr on one example and does not prove that pr(i) is reduced or that pr preserves the crystal structure.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Unveiling Crystal Embeddings: New Perspectives on String Polytopes and Atomic Decompositions." pith.science (2026). https://pith.science/paper/LFO7NII5

@misc{pith2026250522127,
  author       = {Pith},
  title        = {Pith review of: Unveiling Crystal Embeddings: New Perspectives on String Polytopes and Atomic Decompositions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LFO7NII5}},
  note         = {Machine review of arXiv:2505.22127}
}

read the original abstract

We present n-1 different embeddings of string polytopes of type A. We characterize their compatibility with the crystal structure on the string polytopes, and formulate a conjecture describing how to obtain n-1 different atomic decompositions of the crystal with highest weight a multiple of the highest root.

Figures

Figures reproduced from arXiv: 2505.22127 by the authors.

**Figure 1.** The wiring diagram wd(i) for n = 5 and i = (1, 2, 3, 4, 1, 2, 3, 1, 2, 1) with orientation (ℓ3, ℓ4). In blue, a rigorous path in Γ3 determined by its turning points v3,1 and v1,4. 3.1. Rigorous paths. In this subsection we recall the combinatorics of rigorous paths and how they encode inequalities for string cones following Gleizer and Postnikov [10] and crystal operators following Genz, Koshevoy and Schumann [9]. … view at source ↗

**Figure 2.** On the left (resp. right) the two red arrows are forbidden in left (resp. right) rigorous paths. Remark 3.3. Rigurous paths appear as Reineke crossings in [9]. Rigorous paths allow us to describe the string cones and polytopes comnbinatorially as follows. Theorem 3.4 ([10] Corollary 5.8, [9] Theorem 6.1). Let i be a reduced expression of w0. Then the corresponding string cone is given by Si = {x ∈ R N : ⟨x, r(γ)⟩ … view at source ↗

**Figure 3.** Half of the associated crystal graph B(θ) for sl(5, C) and its adjoint representation with highest root θ = ω1 + ω4. Arrows a i −→ b mean fi(a) = b. ii. For each 1 ≤ j ≤ n − 1, the coordinates of zn−j in Si are given by greedily finding in i the word n − j · · ·ij (with ij = n − 1 if j = n − 1, and ij = 1 otherwise) as a subword (up to commutation), reading everything from left to right, placing 1’s in these coordi… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Notice how the subword of i = (1, 3, 2, 4, 1, 3, 2, 4, 3, 1) corresponding to z1 = f1f2f3f4.bθ (resp. z4 = f4f3f2f1.bθ) picks up all crossings along the wire ℓ1 (resp. ℓ5). This is not true for other zi , i ̸∈ {1, 4}, as for example, z3 = f3f4f2f1.bθ is parametrized by…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages

[1]

Berenstein and A

A. Berenstein and A. Zelevinsky. Tensor product multiplicities, canonical bases and totally positive varieties.Inventiones Mathematicae, 143(2):77–128, 2001

work page 2001
[2]

Bossinger and G

L. Bossinger and G. Fourier. String cone and superpotential combinatorics for flag and Schu- bert varieties in type A.J. Combin. Theory Ser. A, 167:213–256, 2019

work page 2019
[3]

Bump and A

D. Bump and A. Schilling.Crystal Bases. Representations and Combinatorics.World Scien- tific Publishing Co. Pte. Ltd., 2017

work page 2017
[4]

L. M. Butler.Subgroup lattices and symmetric functions, volume 539. American Mathemat- ical Soc., 1994

work page 1994
[5]

P. Caldero. Toric degenerations of Schubert varieties.Transform. Groups, 7(1):51–60, 2002

work page 2002
[6]

T. S. Developers.SageMath. SageMath, the Sage Mathematics Software System (Version 9.2), 2020

work page 2020
[7]

X. Fang, G. Fourier, and P. Littelmann. Essential bases and toric degenerations arising from birational sequences.Adv. Math., 312:107–149, 2017

work page 2017
[8]

V. Genz, G. Koshevoy, and B. Schumann. Polyhedral parametrizations of canonical bases & cluster duality.Adv. Math., 369:107178, 41, 2020

work page 2020

Show all 22 references

[9]

V. Genz, G. Koshevoy, and B. Schumann. Combinatorics of canonical bases revisited: string data in type A.Transformation Groups, pages 1–29, 2022

2022
[10]

Gleizer and A

O. Gleizer and A. Postnikov. Littlewood-Richardson coefficients via Yang-Baxter equation. Internat. Math. Res. Notices, 14:741–774, 2000

2000
[11]

Kashiwara

M. Kashiwara. Crystalizing the q-analogue of universal enveloping algebras.Communications in Mathematical Physics, 133:249–260, 1990

1990
[12]

Kashiwara

M. Kashiwara. On crystal bases of the q-analogue of universal enveloping algebras.Duke Mathematical Journal, 63(2):465–516, 1991. 12 LARA BOSSINGER, JACINTA TORRES

1991
[13]

Kashiwara

M. Kashiwara. Crystal bases of modified quantized enveloping algebra.Duke Mathematical Journal, 73(2):383–413, 1994

1994
[14]

Kashiwara

M. Kashiwara. On crystal bases. InRepresentations of groups (Banff, AB, 1994), volume 16 ofCMS Conf. Proc., pages 155–197. Amer. Math. Soc., Providence, RI, 1995

1994
[15]

Lascoux and M.-P

A. Lascoux and M.-P. Sch¨ utzenberger. Sur une conjecture de ho foulkes.CR Acad. Sci. Paris S´ er. AB, 286(7):A323–A324, 1978

1978
[16]

Lecouvey and C

C. Lecouvey and C. Lenart. Atomic decomposition of characters and crystals.Adv. Math., 276(51), 2021

2021
[17]

Libedinsky, L

N. Libedinsky, L. Patimo, and D. Plaza. Pre-canonical bases on affine Hecke algebras.Adv. Math., 399:Paper No. 108255, 2022

2022
[18]

Littelmann

P. Littelmann. Cones, crystals, and patterns.Transform. Groups, 3(2):145–179, 1998

1998
[19]

G. Lusztig. Canonical bases arising from quantized enveloping algebras.J. Amer. Math. Soc., 3(2):447–498, 1990

1990
[20]

G. Lusztig. Finite-dimensional Hopf algebras arising from quantized universal enveloping algebra.J. Amer. Math. Soc., 3(1):257–296, 1990

1990
[21]

L. Patimo. Charges via the affine Grassmannian.preprint arXiv: 2106.02564, 2021

2021 arXiv
[22]

Patimo and J

L. Patimo and J. Torres. Atoms and Charge in typeC 2.Algebraic Combinatorics, 8(2):521– 574, 2025

2025

Pith tools

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

[1] [1]

Berenstein and A

A. Berenstein and A. Zelevinsky. Tensor product multiplicities, canonical bases and totally positive varieties.Inventiones Mathematicae, 143(2):77–128, 2001

work page 2001

[2] [2]

Bossinger and G

L. Bossinger and G. Fourier. String cone and superpotential combinatorics for flag and Schu- bert varieties in type A.J. Combin. Theory Ser. A, 167:213–256, 2019

work page 2019

[3] [3]

Bump and A

D. Bump and A. Schilling.Crystal Bases. Representations and Combinatorics.World Scien- tific Publishing Co. Pte. Ltd., 2017

work page 2017

[4] [4]

L. M. Butler.Subgroup lattices and symmetric functions, volume 539. American Mathemat- ical Soc., 1994

work page 1994

[5] [5]

P. Caldero. Toric degenerations of Schubert varieties.Transform. Groups, 7(1):51–60, 2002

work page 2002

[6] [6]

T. S. Developers.SageMath. SageMath, the Sage Mathematics Software System (Version 9.2), 2020

work page 2020

[7] [7]

X. Fang, G. Fourier, and P. Littelmann. Essential bases and toric degenerations arising from birational sequences.Adv. Math., 312:107–149, 2017

work page 2017

[8] [8]

V. Genz, G. Koshevoy, and B. Schumann. Polyhedral parametrizations of canonical bases & cluster duality.Adv. Math., 369:107178, 41, 2020

work page 2020