MathAudit

Paper metadata

Original submission
March 25, 2026
Claimed MSC
arXiv categories
math.CO
Comments
23 pages, 9 figures

Abstract

We consider real hyperplane arrangements whose hyperplanes are of the form {xi−xj=s}\{x_i-x_j=s\} for some integer ss, which we call deformations of the braid arrangement. In 2018, Bernardi gave a counting formula for the number of regions of any deformation of the braid arrangement A\mathcal{A} as a signed sum over some decorated trees. He further showed that each of these decorated trees can be associated to a region RR of the arrangement A\mathcal{A}, and hence we can consider the contribution of each region to the signed sum. Bernardi also implicitly showed that for transitive arrangements, the contribution of any region of the arrangement is 11. We remove the transitivity condition, showing that for any deformation of the braid arrangement the contribution of a region to the signed sum is 11. This provides an alternative proof of the original counting formula, and sheds light on the geometry underlying the formula. We further use this new geometric understanding to better understand the contribution of a tree.

Dependence graphs

AI-generated audit

Audit summary

Audited against arXiv v1

Not a correctness certificate. A “Correct” result may include yellow typos or minor formal corrections that do not affect substantive soundness. It means this audit found no unresolved substantive error under the stated criteria; it does not replace expert scrutiny or formal verification.

Current report

Detailed mathematical audit

Generated August 23, 2026
01Statements4 reported findingsContains wrong statements

The central contribution-one theorem and the final factorization of a tree contribution are correct after the verified proof repairs recorded below. One ancillary assertion is false: a minimal circular interval family containing a singleton need not have at least two connected components. The statement is unused and is repaired by replacing t≥2t\ge2 with t≥1t\ge1. The all-empty arrangement also needs the convention m=0m=0 because its displayed maximum is otherwise undefined.

Theorem 3.1Correct after verified proof repairs

Every region has signed contribution one

PDF pages 6–7 and 15–16 · Theorem 3.1 and proof · arXiv:2603.24885v1

The marked-tree face bijection restricts to the SS-boxed trees attached to a fixed region, and the number of boxes is the dimension of the corresponding relatively open face. Those faces form a finite cell decomposition of the region. Using compactly supported Euler characteristic gives ∑F(−1)n−dim⁡F=(−1)nχc(R)=(−1)n(−1)n=1,\sum_F(-1)^{n-\dim F}=(-1)^n\chi_c(R)=(-1)^n(-1)^n=1, because the open convex region RR is homeomorphic to Rn\mathbb R^n. The repaired Lemma 3.17 and this sign convention prove the theorem for every nonempty arrangement datum; adopting m=0m=0 covers the empty datum as well.

Exact arXiv version 1 ↗
Proposition 4.19Correct after verified local completions

The circular-component factorization of a tree contribution is correct

PDF pages 21–22 · Proposition 4.19 and proof · arXiv:2603.24885v1

After identifying incident faces with the proper subsets of the affine-alcove facet set, the forbidden interval conditions separate over the circular connected components JiJ_i. If the complement SS is nonempty, summing (−1)∣D∗∣(-1)^{|D^*|} over its subsets vanishes; if it is empty, the remaining independent sums multiply to ∏iw(Ji)\prod_i w(J_i). This is exactly the displayed dichotomy. The proof remains valid when there is only one component.

Paragraph after Example 4.15Incorrect · verified local repair

A singleton interval does not force two circular components

PDF page 20 · paragraph immediately before Definition 4.16 · arXiv:2603.24885v1

The manuscript infers t≥2t\ge2 solely because the inclusion-minimal interval family IS\mathcal I_S contains a singleton. This implication is false: IS={{1}}\mathcal I_S=\{\{1\}\} has one circular connected component. The configuration can also occur in the paper's geometric setting; for n=2n=2, take a deformation whose only hyperplane is x1−x2=−2x_1-x_2=-2 and the 2-Catalan alcove 1<x1−x2<21<x_1-x_2<2. The deformation hyperplane is not face-supporting there, while exactly the outer facet is nonseparating, so the minimal interval family is a singleton. Replace t≥2t\ge2 by t≥1t\ge1. No later argument uses the stronger inequality.

Exact arXiv version 1 ↗
Definitions 2.6 and 3.1Minor scope defect · verified repair

The empty arrangement needs a maximum convention

PDF pages 5–6 · Definition 2.6 and Theorem 3.1 · arXiv:2603.24885v1

The theorem allows every collection of finite sets Sa,bS_{a,b}, including the case in which all of them are empty, but then m=max⁡{∣s∣:s∈Sa,b}m=\max\{|s|:s\in S_{a,b}\} is undefined. Set m=0m=0 when the union of all intercept sets is empty. The empty arrangement has one region, its tree sum reduces to the same Euler-characteristic identity, and all subsequent definitions are then meaningful.

02Proofs4 reported findingsContains incorrect or incomplete proofs

The main proof is not valid exactly as printed. Lemma 3.17 confuses edge and path hyperplanes, reverses orientations and indices, and infers absence of a summed hyperplane only from absence of its individual edge hyperplanes. The two Euler-characteristic arguments use ordinary Euler characteristic with the compact-support sign. Section 4 also omits the onto parts of two affine-alcove identifications. All of these defects have direct verified repairs, and the repaired proof establishes the central theorem and final factorization.

Lemma 3.17Incorrect and incomplete as written · verified repair

The restriction to SS-boxed trees needs a path-sum argument

PDF pages 13–15 · proof of Lemma 3.17 · arXiv:2603.24885v1

Several printed steps cannot be used literally. The preliminary zero-child condition says v<0-child(u)v<0\text{-child}(u) while assuming v=0-child(u)v=0\text{-child}(u); it must say u<vu<v. The inverse notation later writes α−1(T,B)\alpha^{-1}(T,B) although the input is the marked tree (T,μ)(T,\mu); it must be α−1(T,μ)=(T,Bμ)\alpha^{-1}(T,\mu)=(T,B_\mu) (and the earlier βμ\beta_\mu label is likewise BμB_\mu). The first implication reverses the path endpoints and its Si,j−S^-_{i,j} orientation. The converse writes xj−xix_j-x_i where the vertices are vj,viv_j,v_i, ranges jj up to nn instead of the path length, and equates an intersection of edge hyperplanes with one hyperplane. Most importantly, from the fact that each marked edge hyperplane is absent from ASA_S it concludes that the summed hyperplane is absent, which does not follow. The repair is direct. For path positions p<qp<q, put t=∑r=pq−1lsib⁡(vr+1)≥0.t=\sum_{r=p}^{q-1}\operatorname{lsib}(v_{r+1})\ge0. The marked equalities imply containment of the face in {xvq−xvp=t}\{x_{v_q}-x_{v_p}=t\}. When t=0t=0, the marked-zero rule gives vp<⋯<vqv_p<\cdots<v_q; with that orientation, the definition of Svp,vq−S^-_{v_p,v_q} says precisely that t∈Svp,vq−t\in S^-_{v_p,v_q} iff this summed hyperplane belongs to ASA_S. Because F∈FSF\in\mathcal F_S excludes containment in every hyperplane of ASA_S, the sum is not in S−S^-. Conversely, if F∉FSF\notin\mathcal F_S, orient the containing arrangement hyperplane so its displacement is nonnegative. The inverse marked tree supplies the corresponding marked cadet subpath, whose sum belongs to S−S^-, so its box is not an SS-cadet sequence. This proves exactly βm(US)=FS\beta_m(U_S)=\mathcal F_S; the remaining incidence argument restricts the bijection to each region and the component count gives the dimension.

Exact arXiv version 1 ↗
Corollary 3.10 and proof of Theorem 3.1Incorrect convention as written · verified repair

The finite open-cell sum is a compactly supported Euler characteristic

PDF pages 10–11 and 15–16 · proofs of Corollary 3.10 and Theorem 3.1 · arXiv:2603.24885v1

The region is generally unbounded and is partitioned into relatively open, noncompact cells. Ordinary Euler characteristic is not finitely additive over that partition with the displayed signs. Use compactly supported Euler characteristic. A relatively open dd-cell has χc=(−1)d\chi_c=(-1)^d, and an open convex nn-region has χc=(−1)n\chi_c=(-1)^n. Therefore ∑F(−1)n−dim⁡F=(−1)n∑F(−1)dim⁡F=(−1)nχc(R)=1.\sum_F(-1)^{n-\dim F}=(-1)^n\sum_F(-1)^{\dim F}=(-1)^n\chi_c(R)=1. This fixes both occurrences and preserves every conclusion.

Lemmas 4.6–4.7, Remark 4.8, and Proposition 4.12Incomplete as written · verified repair

The affine-alcove face model needs its missing onto arguments

PDF pages 18–20 · Lemmas 4.6–4.7, Remark 4.8, and Proposition 4.12 · arXiv:2603.24885v1

Lemma 4.6 fixes one supporting hyperplane and then states an equivalence that only holds after quantifying over every hyperplane containing FF. Lemma 4.7 checks that the affine map sends a relatively bounded m-Catalan chamber into the standard alcove but does not prove equality. Relative boundedness rules out every outer interval, so each pairwise difference lies between consecutive integer walls already present in the m-Catalan arrangement; the chamber is therefore one affine type-AA alcove, and the permutation/integer translation maps it onto the standard alcove. Remark 4.8's independence of mm follows by appending only rightmost leaf children: cadet edges, left-sibling counts, and SS-boxings do not change. Finally, every proper subset of the nn standard-alcove facet equations has a unique nonempty relatively open face, while the full set is inconsistent. This supplies the omitted surjectivity of τ\tau in Proposition 4.12 and completes the Boolean-face identification used by Proposition 4.19.

Local notation in Lemma 3.17Typo

The final region restriction uses two stale symbols

PDF page 15 · final paragraphs of Lemma 3.17 · arXiv:2603.24885v1

The general-deformation proof refers to a region of AGA_G, ends with FR\mathcal F_R, and restricts an undefined bare β\beta. These are copied or stale symbols. They must be a region of ASA_S, the set FS(R)\mathcal F_S(R), and the already defined map βS\beta_S, respectively, so the final line is βR=βS∣US(R):US(R)→FS(R)\beta_R=\beta_S|_{U_S(R)}:U_S(R)\to\mathcal F_S(R). The surrounding definitions determine every replacement uniquely.

03Novelty0 reported findingsNo non-novelty findings

No non-novelty findings.

04Sources4 reported findingsSources correct and complete

Every proof-critical external result was traced in its exact cited form. Bernardi's 2025 Theorem 3.8 supplies precisely the marked-tree face bijection used in Theorem 3.12; Bernardi's 2018 region bijection is also used directly and recursively. The 2018 version of record has a real inverse-construction defect, but its corrected arXiv version 4 repairs the exact theorem. Levear's Theorem 14 is an independently sufficient second source for the m-Catalan face bijection, not an additional logical prerequisite. No missing proof-critical citation was found.

Theorem 3.12 via Bernardi 2025Correctly invoked

The marked-tree face theorem matches exactly

Focal PDF pages 12–15 · Theorem 3.12 and Lemma 3.17 · arXiv:2603.24885v1; Bernardi 2025 PDF pages 6 and 21–28 · Theorem 3.8

The m-Catalan arrangement is strongly transitive, so Bernardi's Theorem 3.8 applies. Its marked edges encode equality along cadet edges, its unmarked inequalities are the same Ψ\Psi formula reproduced by the focal paper, and its number of marked edges is codimension. Since a boxing with ∣B∣|B| cadet components has n−∣B∣n-|B| marked edges, the face dimension is exactly ∣B∣|B|, which is the statistic required by Lemma 3.17.

Olivier Bernardi, arXiv:2503.00662v1 ↗
Direct and recursive Bernardi 2018 inputCorrect after verified source repair

The exact region theorem is valid after its documented repair

Focal PDF page 6 and Lemma 3.17 on page 15 · Theorem 2.9 · arXiv:2603.24885v1; Bernardi 2025 PDF pages 3 and 21–22 · Theorem 2.9 and proof of Theorem 3.8; Bernardi 2018 journal pages 503–513

The focal paper uses Bernardi's tree-to-Catalan-region map directly to define TRT_R and to identify the incident unmarked region of a face. Bernardi 2025 also invokes the same theorem on every restricted flat. In the exact 2018 version of record, property (iv)'s fixed increment can create equal consecutive zz-values. ArXiv:1604.06554v4 replaces it by 2−p2^{-p} and changes the second branch to s≠0s\ne0, a repair verified against the downstream Proposition 8.1 and Theorem 8.8 proof. Theorem 8.8's surjectivity is internal; its transitive injectivity uses Theorem 4.6, whose proof-critical published stopping sources are Zaslavsky's Memoirs monograph and Stanley's Enumerative Combinatorics, Volume 2.

Bernardi, corrected arXiv:1604.06554v4 ↗
Levear 2021 alternativeCorrect alternative source

The independently cited Catalan-face bijection also supports Theorem 3.12

Focal PDF page 12 · Theorem 3.12 citation [4, Theorem 14] · arXiv:2603.24885v1; Levear version of record, Theorem 14 and Sections 2.2–3

Levear's exact open version of record proves a dimension-graded bijection between faces of the extended m-Catalan arrangement and decorated (m+1)(m+1)-ary trees. After the notation translation recorded by Bernardi 2025, it gives the same theorem used by the focal paper. Because Bernardi 2025 already supplies the full required statement and the focal proof does not combine separate hypotheses from the two papers, Levear is recorded as an independently sufficient alternative rather than a second mandatory graph edge.

Duncan Levear, Electronic Journal of Combinatorics 28(4) (2021), P4.29 ↗
Remaining bibliographyComplete for the marked proof scope

Contextual and reproduced results are correctly excluded from the proof graph

Focal PDF pages 1–2, 5–8, 16, and 22–23 · introduction, background, Section 4 motivation, and references · arXiv:2603.24885v1

The Bernardi signed-count theorem is recovered as a consequence rather than assumed in proving Theorem 3.1. Bisain–Hanson supplies context and an earlier tree-contribution algorithm but no step of the new proof. Orlik–Terao, Postnikov–Stanley, Stanley's survey, and Zaslavsky's theorem are background citations in the focal manuscript; no unproved step in Lemma 3.17 or Proposition 4.19 requires them directly. Their omission as direct focal edges therefore reflects logical dependence, not missing bibliography.

Detailed audit reportFull reasoning, manuscript locations, and references.
Open report PDF ↗

Author response

Challenge an audit finding

Local workflow preview

A listed author may submit formal evidence that an audit is inaccurate. The response would be considered in a fresh AI re-evaluation; it would not edit the audit automatically.

Paper
arXiv:2603.24885v1
Authors listed
Neha Goregaokar, Aaron Lin
Audit date
August 23, 2026
  1. 01Establish identityMatch an authenticated scholarly identity to this paper.
  2. 02Submit evidenceIdentify the finding and give a formal mathematical response.
  3. 03Re-evaluateA separate agent checks the response and records a disposition.
Recommended production method

Authenticate with ORCID, then require an exact arXiv match

MathAudit should accept the identity only when ORCID OAuth authenticates the claimant's iD and this exact arXiv paper appears in arXiv's public authority feed for that iD. A matching name alone is not sufficient.

ORCID OAuth and arXiv authority-record lookup are not connected in this local prototype.

Email fallback for papers without a linked ORCID

A production fallback could send a one-time link only when the submitted address matches an independently maintained author-contact allowlist for this paper. MathAudit must return the same message for every address so the form cannot reveal which contacts are on that list.

This demonstration does not send, store, or compare the address.

Structured response preview

This form remains unavailable until production identity verification succeeds. Nothing entered here is submitted.

This panel never establishes authorship in the local prototype. A production result should be described narrowly as an authenticated ORCID match or control of a separately allowlisted author-contact mailbox.