On the minimality of pancake decomposition of surface germs

Davi Lopes Medeiros; Emanoel Souza; Euripedes Carvalho da Silva

REVIEW 3 major objections 4 minor 18 references

On the minimality of pancake decomposition of surface germs

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 This paper proves that a greedy word-reading algorithm produces a pancake decomposition with the minimal number of pieces for every snake and circular snake surface germ, and that the decomposition is canonical up to weak outer…

desk verdict A useful, mostly sound algorithmic paper whose greedy construction leans on an unproved word-geometry equivalence that a referee should ask to be made explicit. read the letter →

arxiv 2505.23976 v1 pith:H3NZX7A6 submitted 2025-05-29 math.MG math.AG

Davi Lopes Medeiros , Euripedes Carvalho da Silva , Emanoel Souza This is my paper

classification math.MGmath.AG MSC 14P1003C6451F30

keywords pancakedecompositionsnakecircularsurfacegermsouterLipschitzequivalencenormallyembeddedminimalsequencegreedyalgorithm

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 works on the outer Lipschitz classification of surface singularities, where the basic abnormal pieces are called snakes and circular snakes. It claims that for every such germ there is an algorithmically constructed pancake decomposition with the minimum possible number of pancakes, and that any two weakly outer bi-Lipschitz equivalent snakes or circular snakes have greedy decompositions that are themselves weakly equivalent. The algorithm reads the combinatorial snake name as a word and cuts at the first letter that repeats a previous letter, a step repeated until the word is exhausted. If the theorems are correct, every snake and circular snake carries a canonical minimal decomposition, giving a concrete invariant for the still-open outer classification problem.

What carries the argument

The load-bearing object is the snake name, a word whose letters are the nodal zones of the surface, and the greedy rule that reads it. A word segment is primitive exactly when it has no repeated letter; geometrically this corresponds to a normally embedded piece, so the algorithm cuts a surface wherever the accumulated word first repeats a letter. The resulting minimal (or fundamental) sequence supplies the boundary arcs of the pancakes. For circular snakes with nodal zones, the same rule applied to the infinite periodic word can return to its starting zone after more than one lap, and the paper's lifting construction (Lemma 5.10) bounds the period by the number of laps, reducing the circular problem to the snake problem. The choice of the actual arc inside each segment is constrained by a single horn-neighborhood condition that guarantees each H\"older triangle is LNE.

What would settle it

Run the greedy algorithm on any concrete snake name, for instance $W=[abacdbcd]$, and check that each pair of adjacent pancakes has a non-LNE union and that the number of pancakes cannot be reduced by merging. More decisively, search the finite set of circular snake names for a fundamental sequence $\{j_k\}$ that violates the inequality $p \geq t(q-1)+1$ from Lemma 5.10; the theory says none exists.

Watch

Extended reading notes

Core claim

The central discovery is that minimality for snakes and circular snakes is a word-combinatorial phenomenon, not a metric search. For a snake, the greedy cut points are the entries of the minimal sequence $j_0=0$, $j_i=\min\{k>j_{i-1} : [x_{j_{i-1}}\cdots x_k] \text{ is not primitive}\}$, and the pieces $T(\lambda_{i-1},\lambda_i)$ form a minimal pancake decomposition (Theorem 4.9). The same rule, adapted to the eventually periodic sequence of a circular snake with nodal zones, yields a minimal decomposition (Theorem 5.11), while a circular snake without nodal zones of multiplicity $m$ yields $m+1$ pieces (Theorem 5.3). Propositions 6.1 to 6.3 then show that weak outer bi-Lipschitz equivalence preserves the greedy decomposition, so the minimal number of pancakes and the pattern of cut points are canonical weak invariants. The paper also proves this canonicity is sharp: it fails for the stronger outer equivalence, and the greedy rule need not give minimal decompositions for arbitrary H\"older triangles.

Load-bearing premise

The construction assumes that a segment of the snake name with no repeated letter exactly matches the geometric fact that the corresponding part of the surface is Lipschitz normally embedded, and that this dictionary between words and geometry has no exceptions beyond the snakes and circular snakes already classified.

Editorial extensions

If this is right

Every snake and every circular snake has a computable pancake decomposition of provably minimal size.
The number of pancakes in the greedy decomposition is a weak outer bi-Lipschitz invariant, shared by all weakly equivalent surfaces.
The greedy boundary arcs can be chosen inside segments with a prescribed horn-neighborhood condition, giving explicit cut points rather than an existence argument.
The word-level description makes minimal decompositions checkable by combinatorial computation: any weakly outer bi-Lipschitz map between equivalent snakes must send the greedy cut pattern to the corresponding pattern.
The sharpness examples show that the same canonicity fails under full outer equivalence, so weak equivalence is exactly the level where the decomposition is canonical.

Reading between the lines

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

One testable extension is an exhaustive computer search over snake names and circular snake names of small length, verifying that the greedy cut points always produce reduced decompositions and that the number of pieces equals the minimal count; this would confirm the theorems in a finite range.
The inequality $p \geq t(q-1)+1$ from Lemma 5.10 may be a purely word-combinatorial statement; if so, it could be detached from surface geometry and checked or generalized in symbolic dynamics.
Since any surface germ is either a circular snake or contains finitely many snakes, the greedy decomposition gives a canonical candidate for the abnormal part of an arbitrary germ; the failure examples for H\"older triangles suggest that a complete outer classification will need to record tangency orders within segments, not just the letter multiplicities.

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

Punchline: the paper does something new and useful—gives a greedy algorithm for minimal pancake decompositions of snakes and circular snakes and proves those decompositions are canonical up to weak outer bi-Lipschitz equivalence—but one load-bearing equivalence is asserted without proof, and the referee should ask for it before signing off.

What is actually new: minimality of the greedy decompositions (Theorems 4.9, 5.3, 5.11), the fundamental-vs-minimal sequence distinction for circular snakes (Definition 5.5, Example 5.9, Lemma 5.10), and the weak canonicity results (Propositions 6.1–6.3), plus the sharpness examples showing that outer canonicity fails (Example 7.3) and that a naive 'join LNE adjacent pieces' procedure does not give a minimal decomposition in general (Example 7.2). The exposition is careful about what is being assumed from [13] and [10]. I credit the algorithmic clarity: the greedy construction is concrete and the examples (Figures 4, 5, 8, 9) are genuinely helpful.

The soft spot is real. Remarks 4.3 and 5.7 state, as a 'direct consequence' of the geometric Definition 4.1, that a Holder triangle T(θ_a, θ_b) between nodal zones fails to be LNE exactly when the corresponding snake-name subword [x_a...x_b] is not primitive. No proof or citation is given. This equivalence is not decorative: the greedy cut points j_i are defined as the first place this condition holds. If the word-level condition has any exception—say, a failure of LNE caused by cluster multiplicities that are not visible in the word before the first repeated letter—then the algorithm is computing a different sequence from the geometric minimal sequence, and the lower-bound argument in Theorem 4.9 (and its circular analogues) does not go through. This needs a short proof or an explicit pointer to a lemma in [13] or [10].

The other soft spots are milder. Theorem 4.9 says the proof for i>2 'follows analogously', and Theorem 5.11 and Proposition 6.1 have similar compressed passages, including the bubble/spiral-snake case at the end of Proposition 6.1. These are probably routine, but in a paper whose main claim is an algorithm, 'follows analogously' is exactly where the reader needs to see the induction or the case split. The reliance on the authors' own classification theorems is not circular, but it concentrates the correctness risk.

Bottom line: this is a serious paper for the Lipschitz geometry of surface germs. I would send it to peer review and ask for the missing proof of the word-geometry equivalence and an expansion of the analogous cases. If those come back clean, it is a solid, citable contribution.

Referee Report

3 major / 4 minor

Summary. The paper studies the outer Lipschitz geometry of 'snakes' and 'circular snakes', abnormal surface germs introduced by Gabrielov and Souza. It defines a 'greedy pancake decomposition' via a minimal sequence of cut points computed from the snake name, and proves three main results: the greedy decomposition is a minimal pancake decomposition for snakes (Theorem 4.9) and for circular snakes with and without nodal zones (Theorems 5.3 and 5.11), and greedy decompositions of weakly outer equivalent snakes or circular snakes are themselves weakly outer bi-Lipschitz equivalent (Propositions 6.1–6.3). The paper also includes examples showing that the greedy algorithm fails for general Hölder triangles (Example 7.2) and that the canonicity does not hold for the stronger notion of outer equivalence (Example 7.3).

Significance. If correct, the paper gives the first constructive, canonical minimal pancake decompositions for snakes and circular snakes, which are key objects in the outer Lipschitz classification problem. The algorithms are explicit and the examples are informative, and the paper properly credits the classification theorems of [13] and [10] on which the canonicity results depend. The main reservation is that the reduction from the geometric non-LNE condition to a combinatorial non-primitivity condition is asserted without proof; since this reduction is the engine of the greedy algorithm, the central claims are not yet fully established.

major comments (3)

[§4 Remark 4.3 and §5 Remark 5.7] The equivalence between the geometric condition 'T(θ_a, θ_b) is not LNE' and the combinatorial condition 'the subword [x_a … x_b] contains a repeated letter' is asserted as a 'direct consequence' of Definitions 4.1 and 5.5, but no proof or supporting reference is given. This equivalence is load-bearing: the greedy cut points j_i are exactly the first positions where the equivalence is invoked, and the minimality lower bound in Theorem 4.9 (and its circular analogues) depends on those cut points being precisely the pairs N_{j_{i-1}}, N_{j_i} that cannot lie in a common LNE pancake. If the equivalence has any exception—for example, a sub-snake whose failure of LNE is caused by cluster multiplicities or node data not visible in the word before the first repeated letter—the algorithm may produce different cut points and the lower-bound argument would not apply. The paper needs a standalone proof of this equivalence or an exact citation to a lemma in [13] or [10] that establishes it.
[§4 Remark 4.2 and §5 Remark 5.6] The manuscript asserts that the minimal sequence is independent of the choice of the arcs θ_i in the nodal zones, but no proof is provided. This independence is essential for the well-definedness of the greedy decomposition (Definition 4.10) and for the canonicity statements in Propositions 6.1–6.3. The assertion may be derived from the word-level characterization of Remark 4.3 once that characterization is proved, but as written it is an unproved claim about the geometric definition. Please supply a proof or a precise reference.
[§4 Theorem 4.9; §5 Theorem 5.11; §6 Proposition 6.1] Several key steps are deferred with 'follows analogously' or 'admits the same adaptation': the proof that X_i is LNE for i > 2 in Theorem 4.9, the corresponding part of Theorem 5.11, and the treatment of bubbles and spiral snakes in Proposition 6.1. The case tord(S_{j_2}, S_{j_1}) > β in the proof of Theorem 4.9 shows that the choice of λ_i in condition (2) is genuinely delicate, so the remaining cases are not identical to the ones written out. The reader cannot verify without an explicit induction or a clear statement of why the analogous argument applies to every i. Please expand these passages enough to make the inductive step transparent, and spell out the adaptation for bubbles and spiral snakes in Proposition 6.1.

minor comments (4)

[§5 Theorem 5.3] In the proof of Theorem 5.3 the notation {X_i}_{p+1}^{i=1} is used with p undefined; it should be {X_i}_{m+1}^{i=1} throughout the proof.
[§2 Theorem 2.32] The theorem is stated as '(Theorem 8.3 in [10])', but the statement covers β-snakes as well as circular β-snakes; the snake case is Theorem 6.28 of [13], not Theorem 8.3 of [10]. Please correct the attribution or split the statement according to the source.
[§6 Proposition 6.1] In the proof of Proposition 6.1, the sentence 'θ'_0 = γ'_1, θ'_n = γ_2' should read 'θ'_0 = γ'_1, θ'_n = γ'_2'; the current wording is a typo.
[Throughout] There are several minor typographical issues, including 'lenght' for 'length' in Remark 2.27, 'bubles' for 'bubbles' in Example 7.1, and 'the the desired weakly outer homeomorphism' in the proof of Proposition 6.1. A careful proofreading pass is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: greedy minimality is proved from the geometric non-LNE breakpoints, not assumed; the word-level shortcut and self-cited classification are external inputs, not definitional reductions.

full rationale

The minimality arguments are not circular. In Definition 4.1 the greedy/minimal sequence is defined geometrically, by the first indices j_i for which the Hölder triangle T(θ_{j_{i-1}}, θ_{j_i}) is not LNE. Theorem 4.9 then proves that the resulting decomposition is a pancake decomposition and that any pancake decomposition needs at least p+1 pancakes, precisely because each pair N_{j_{i-1}}, N_{j_i} cannot lie inside a single LNE pancake. This lower bound is derived from the defining non-LNE property, not from assuming the conclusion. The analogous proofs for circular snakes (Theorems 5.3 and 5.11) are similarly built from the non-LNE pairs produced by the fundamental/minimal sequences or from the multiplicity count, rather than from the desired conclusion. The word-based computation in Remarks 4.3 and 5.7 is stated as a 'direct consequence' of the geometric definition and is not proved in detail; this is a legitimate correctness risk if the asserted equivalence between non-LNE triangles and non-primitive subwords has exceptions, but it is not a circular reduction, because the geometric minimal sequence is defined independently and the minimality proof does not invoke the word-based version. The canonicity propositions in Section 6 do rely on the weak classification theorems quoted from [13] and [10], including work by overlapping authors; this is a self-citation with real load-bearing weight for the canonicity claim, but it functions as an external classification input rather than as a restatement of the greedy construction, and the core minimality result is not defined in terms of that classification. Finally, Example 7.2 shows that the greedy algorithm can fail for general Hölder triangles, confirming that the minimality step carries genuine mathematical content rather than being a tautology. No fitted parameter is relabeled as a prediction, and no uniqueness theorem is invoked to force the greedy choice by definition.

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

The paper's claims rest on the definable o-minimal framework and on the classification machinery of abnormal surface germs built in [13] and [10]. The present text supplies the greedy algorithm, minimality proofs, and canonicity proofs, but does not reprove the external classification theorems. No numerical fitting or free parameters are involved. The most delicate axioms are the word-level equivalence and the lifting and realization theorem for circular snakes.

assumptions (6)

domain assumption All sets and maps are definable in a polynomially bounded o-minimal structure over R with field of exponents F.
Stated at the start of Section 2; it guarantees finiteness, arc selection, and the existence of exponents used throughout.
standard math Arc Selection Lemma and non-archimedean property of tangency order.
Used in Remark 2.5 and in the proof of Theorem 4.9 to choose generic arcs converging to boundary arcs; cited to [3], [7], and [11].
standard math Birbrair's inner classification: connected surface germs are inner bi-Lipschitz equivalent to beta-Holder triangles or beta-horns.
Basis for Definitions 2.7, 2.23, and for the pancake structure; from [1] and [2].
standard math Kurdyka-Orro pancake decomposition existence: every definable set admits a finite pancake decomposition.
Used in Remark 3.5 to assert minimal decompositions exist; from [9], [14], and [15].
domain assumption Structural lemmas for snakes: zones, segments, nodal zones, Proposition 4.27, Proposition 4.30, Proposition 4.56, and Proposition 4.59 of [13].
Used in Lemmas 4.6 and 4.7 and in Proposition 6.1; [13] includes a coauthor of the present paper.
domain assumption Circular snake classification and realization theorems, Theorems 8.3, 8.4, and 7.10 in [10], presented as Theorems 2.32 and 2.33 here.
Load-bearing for Sections 5.2 and 6, especially for the lifting construction in Theorem 5.11; [10] includes two of the present authors.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On the minimality of pancake decomposition of surface germs." pith.science (2026). https://pith.science/paper/H3NZX7A6

@misc{pith2026250523976,
  author       = {Pith},
  title        = {Pith review of: On the minimality of pancake decomposition of surface germs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/H3NZX7A6}},
  note         = {Machine review of arXiv:2505.23976}
}

read the original abstract

The abnormal surfaces called snakes and circular snakes, defined in \cite{GabrielovSouza}, are special types of surface germs capturing the outer Lipschitz phenomena relevant to the outer classification problem. We provide algorithms to obtain a minimal pancake decomposition, i.e., where the number of pancakes is minimal, for snakes and circular snakes. We call a pancake decomposition obtained from our algorithm a greedy pancake decomposition. We also prove that greedy pancake decompositions of weakly outer Lipschitz equivalent snakes (or circular snakes) are weakly equivalent, in the sense that there is a weakly outer bi-Lipschitz homeomorphism between the surfaces mapping each greedy pancake to a greedy pancake. This implies that such minimal decompositions are also canonical up to weakly outer bi-Lipschitz equivalence.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages

[13]

Gabrielov, E

A. Gabrielov, E. F. Souza. Lipschitz geometry and combinatorics of abnormal surface g erms. Selecta Mathematica, New Series, 28:1 (2022). DOI:10.1007/s00029-021-00716-4

work page doi:10.1007/s00029-021-00716-4 2022
[10]

Lipschitz geometry and combinatorics of circular snakes

A. Costa, D.L. Medeiros, E. Souza. Lipschitz geometry and combinatorics of circular snakes. arXiv:2312.04446 (2025)

work page Pith review arXiv 2025
[1]

Birbrair

L. Birbrair. Local bi-Lipschitz classiﬁcation of 2-dimensional semial gebraic sets. Houston J. Math., 25 (1999), no. 3, 453–472

work page 1999
[2]

Birbrair

L. Birbrair. Lipschitz geometry of curves and surfaces deﬁnable in o-min imal structures. Illinois J. Math. V ol. 52 (2008), no. 4, 1325–1353

work page 2008
[3]

Birbrair, A

L. Birbrair, A. Fernandes. Metric theory of semialgebraic curves. Rev. Mat. Complutense, 13 (2000), no. 2, 369-382

work page 2000
[4]

Birbrair, A

L. Birbrair, A. Fernandes, A. Gabrielov, V . Grandjean. Lipschitz contact equivalence of function germs in R2. Annali SNS Pisa, 17 (2017), 81–92, DOI 10.2422/2036-2145.201503 014

work page arXiv 2017
[5]

and Gabrielov, A

Birbrair, L. and Gabrielov, A. Lipschitz geometry of real semialgebraic surfaces. In: Cisneros-Molina, J.L., Dung Tr´ ang, L., Seade, J. (eds). Handbook of Geometry and Topology of Sin gularities IV . Cham: Springer International Publishing, https://doi.org/10.1007/978-3-031-31925-9-8), 449-462 (2023)

work page doi:10.1007/978-3-031-31925-9-8 2023
[6]

Birbrair, A

L. Birbrair, A. Gabrielov. Outer Lipschitz classiﬁcation of normal pairs of H¨ older triangles.” Journal of the London Mathematical Society 111.5: e70168 (2025)

work page 2025

Show all 18 references

[7]

Birbrair, R

L. Birbrair, R. Mendes. Arc criterion of normal embedding. Springer Proceedings in Mathematics & Statistics. V ol. 222 (2018) no. 2015

2018
[8]

Birbrair, R

L. Birbrair, R. Mendes. Lipschitz contact equivalence and real analytic functions . Preprint arXiv:1801.05842 (2018)

2018 arXiv
[9]

Birbrair, T

L. Birbrair, T. Mostowski. Normal embeddings of semialgebraic sets. Michigan Math. J., 47 (2000), 125–132

2000
[11]

van den Dries

L. van den Dries. Tame topology and o-minimal structures Cambridge University Press, (1998)

1998
[12]

van den Dries and P

L. van den Dries and P . Speissegger. O-minimal preparation theorems. Model theory and applications. 87116, Quad. Mat. Aracne, Rome, 11 (2002)

2002
[14]

K. Kurdyka. On a subanalytic stratiﬁcation satisfying a Whitney proper ty with exponent 1. Real algebraic geometry, Springer, (1992), 316–322

1992
[15]

Kurdyka, P

K. Kurdyka, P . Orro. Distance g´ eod´ esique sur un sous-analytique.Revista Math. Univ. Comput. Madrid, 10 (1997)

1997
[16]

Mostowski

T. Mostowski. Lipschitz equisingularity Dissertationes Math. (Rozprawy Mat.), 243 (1985) 46 pp

1985
[17]

Parusinski

A. Parusinski. Lipschitz stratiﬁcations of subanalytic sets , Ann. Sci. Ecole. Norm. Sup., (4) 27 (1994), 661–696

1994
[18]

V alette

G. V alette. The link of the germ of a semi-algebraic metric space. Proc. Amer. Math. Soc., 135 (2007), no. 10, 3083–3090. †DEPARTAMENTO DE MATEM ´ATICA , U NIVERSIDADE FEDERAL DO CEAR ´A (UFC), C AMPUS DO PICI , B LOCO 914, C EP. 60455- 760, F ORTALEZA -C E, BRASIL †DEPARTAME...

2007

Pith tools

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

[1] [13]

Gabrielov, E

A. Gabrielov, E. F. Souza. Lipschitz geometry and combinatorics of abnormal surface g erms. Selecta Mathematica, New Series, 28:1 (2022). DOI:10.1007/s00029-021-00716-4

work page doi:10.1007/s00029-021-00716-4 2022

[2] [10]

Lipschitz geometry and combinatorics of circular snakes

A. Costa, D.L. Medeiros, E. Souza. Lipschitz geometry and combinatorics of circular snakes. arXiv:2312.04446 (2025)

work page Pith review arXiv 2025

[3] [1]

Birbrair

L. Birbrair. Local bi-Lipschitz classiﬁcation of 2-dimensional semial gebraic sets. Houston J. Math., 25 (1999), no. 3, 453–472

work page 1999

[4] [2]

Birbrair

L. Birbrair. Lipschitz geometry of curves and surfaces deﬁnable in o-min imal structures. Illinois J. Math. V ol. 52 (2008), no. 4, 1325–1353

work page 2008

[5] [3]

Birbrair, A

L. Birbrair, A. Fernandes. Metric theory of semialgebraic curves. Rev. Mat. Complutense, 13 (2000), no. 2, 369-382

work page 2000

[6] [4]

Birbrair, A

L. Birbrair, A. Fernandes, A. Gabrielov, V . Grandjean. Lipschitz contact equivalence of function germs in R2. Annali SNS Pisa, 17 (2017), 81–92, DOI 10.2422/2036-2145.201503 014

work page arXiv 2017

[7] [5]

and Gabrielov, A

Birbrair, L. and Gabrielov, A. Lipschitz geometry of real semialgebraic surfaces. In: Cisneros-Molina, J.L., Dung Tr´ ang, L., Seade, J. (eds). Handbook of Geometry and Topology of Sin gularities IV . Cham: Springer International Publishing, https://doi.org/10.1007/978-3-031-31925-9-8), 449-462 (2023)

work page doi:10.1007/978-3-031-31925-9-8 2023

[8] [6]

Birbrair, A

L. Birbrair, A. Gabrielov. Outer Lipschitz classiﬁcation of normal pairs of H¨ older triangles.” Journal of the London Mathematical Society 111.5: e70168 (2025)

work page 2025