A Sharp Four-Layer Theorem for Integer-Occupied Slices of a Planar Disk-Slab

New Energy Technology; Tsinghua University); Xinyue Liu (Institute of Nuclear

REVIEW 3 major objections 3 minor 11 references

A Sharp Four-Layer Theorem for Integer-Occupied Slices of a Planar Disk-Slab

T0 review · 3 major / 3 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper proves that every admissible planar disk-slab instance in the strict-failure branch has at most four occupied integer levels of the reduction-selected functional, and a rational instance attains four.

desk verdict Genuinely new sharp four-layer constant for a narrow but well-defined branch, with a clean proof structure; just make the 17,640-coefficient Bernstein certificate actually available before publication. read the letter →

arxiv 2608.12204 v1 pith:A7PPH6RT submitted 2026-08-12 math.OC

Xinyue Liu (Institute of Nuclear , New Energy Technology , Tsinghua University) This is my paper · ORCID

classification math.OC MSC 11H0690C10

keywords latticeenumerationarithmeticwidthBabainearestplaneplanarconvexbodiesBernsteincertificatesintegerprogrammingreductionsharpbound

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 asks how many integer levels of a deterministically selected direction can contain lattice points when a rational ellipse on an affine lattice coset is cut by a slab. In the nondegenerate strict-failure branch, where the Babai nearest-plane point lies outside the inner ellipse, it proves that the answer is at most four, and it exhibits a rational instance with exactly four occupied levels. The proof rules out every possible block of five consecutive occupied levels, so the bound is exact rather than an artifact of the particular example. A reader should care because the constant four is uniform: it does not grow with the size of the disk or slab once the reduction branch is fixed.

What carries the argument

The carrying object is the reduction-selected primitive-dual functional $w = U^{-T}u_2$ together with the inner-ellipse gauge $H(t,j) = \alpha(t-e_1+\mu j)^2 + \beta(j-e_2)^2$. A deterministic $\delta = 3/4$ reduction fixes the parameters $\alpha,\beta,\mu$ and the half-open error cells; strict Babai failure gives $E_B > 1$ and forces $\beta > 4/3$, which limits occupation to the seven levels $\{-3,\dots,3\}$. A strict diameter estimate $\operatorname{diam}(\tilde C) < 4\sqrt2$ then bounds the span of occupied levels, a nearest-integer inequality kills the central block, and cell localization reduces the two side blocks to eight modes. Four of those modes are excluded by the shared-cap inequality; the remaining four require proving positivity of a four-variable polynomial, which is done by converting its power coefficients to Bernstein coefficients on a rational box and verifying that all 17,640 coefficients are positive.

What would settle it

Recompute the Bernstein coefficient table for the four hard modes in independent exact rational arithmetic and check whether every entry is strictly positive; a single nonpositive coefficient would invalidate Proposition 5.2 and open the possibility of a five-level counterexample.

Watch

Extended reading notes

Core claim

The central claim is $L^*_Z = 4$: for every admissible instance satisfying the strict-failure assumptions, the set $J_Z(I)$ of occupied integer levels of the reduction-induced primitive-dual functional $w = U^{-T} u_2$ has size at most four, and there is a rational admissible instance with $J_Z = \{-2,-1,0,1\}$. The proof first shows that at most seven candidate levels can be occupied, then uses a strict disk-slab diameter estimate to compress any hypothetical five-level occupation into one of three consecutive blocks. A nearest-integer inequality excludes the central block, while layer reversal and cell localization reduce the side blocks to eight explicit modes, of which four violate a shared-cap inequality and four are excluded by an exact Bernstein certificate whose 17,640 coefficients are all positive. Corollary 7.3 then sharpens the uniform bound to $L^*_Z = 4$.

Load-bearing premise

The proof's hard-mode exclusion rests on the claim that all 17,640 exact rational Bernstein coefficients in the certificate are positive; if even one coefficient were nonpositive, the exclusion of the four hard modes would not follow.

Editorial extensions

If this is right

No admissible instance in the strict-failure branch can occupy five or more levels; the only possible five-level blocks are the three listed in Proposition 3.2, and each is excluded.
The rational four-level fixture shows the bound cannot be improved, and in that fixture the real body still intersects five levels, so integer occupation and real layer intersection are genuinely different counts.
Because the safe candidate list is fixed to seven levels and the diameter estimate is uniform, the worst-case count four holds across all sizes of the disk and slab once the reduction branch is fixed.
The proof's exclusions are deterministic finite checks, so any search for occupied levels of the selected functional can stop once four levels have been found.

Reading between the lines

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

Beyond the paper, the Bernstein-certificate exclusion is a finite exact-rational check that could be reused for other eight-mode reductions with the same cap structure, provided the mode-specific polynomials are recomputed.
A testable extension is to vary the deterministic parameters, such as the reduction rounding rule or the value of $\delta$, and search for a five-level instance; finding one would show that the constant four depends on those conventions.
The paper gives no information about higher dimensions; one plausible direction is that the corresponding worst-case number of occupied levels grows with dimension, but the planar proof offers no direct route to such a statement.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Quick take: this is a real result, not a stunt. The sharp four-layer constant L*_Z = 4 for the strict Babai-failure branch is new, and the proof is mostly elementary. The one load-bearing step is a large exact Bernstein certificate that the paper doesn't actually give you, so as printed the theorem is conditional on an unverified computation.

What is actually new: the exact extremal constant, the five-level trichotomy, the eight-mode reduction, and the rational attainment example. The reduction to three block types via the strict diameter estimate is clean, and the nearest-integer inequality for the central block is a nice piece of work. The shared-cap contradiction for the easy modes is also solid and self-contained. The attainment example is fully explicit and checks out arithmetically.

The soft spot is exactly where the reader put their finger: Proposition 5.2 rests on 17,640 positive Bernstein coefficients, and Appendix A only gives the per-mode minima plus a hash digest. That is not enough for a reader to verify the positivity without rerunning the computation themselves. Section 8 says an ancillary program exists, but it isn't in the manuscript. For a computer-assisted proof this is a normal and fixable gap: put the code or the full coefficient table in the ancillary files, or better, make the certificate machine-checkable (e.g., a text file with all coefficients). Until then, Theorem 7.1 is conditional on a computation nobody can see.

There is also a minor arithmetic slip in Section 5: the displayed lower bound for 7(3-vartheta)^2 - 30*vartheta^2 is written as 725/4, but the value at vartheta=1/2 is 145/4. The final conclusion 85/8 is actually correct if you compute the intended expression (-30/4 + (1/2)*(145/4) = 85/8), so this is a typo, not a hole. Still, it shows the exact arithmetic needs a careful pass.

The class is narrow: planar disk-slab, strict failure branch. The paper is honest about that, and the constant is exact for that class. I don't see circularity or hidden fitting; the reduction and Babai failure are fixed algorithmic inputs.

Who's this for: people working on lattice enumeration, fixed-dimension integer programming, or arithmetic width. It gives a sharp worst-case level count for one branch where previously there was no exact constant.

Recommendation: send it to peer review. The mathematical structure is coherent and the result is worth checking. Ask for the certificate/code before final acceptance. If the computation confirms, the paper is publishable as is modulo the typo.

Referee Report

3 major / 3 minor

Summary. The paper studies lattice points in a rational positive-definite quadratic sublevel set intersected with a slab, after a deterministic delta=3/4 reduction selects a primitive-dual integer functional w and an inner ellipse. In the strict-failure branch where the Babai point lies outside that ellipse, the paper claims that the occupied integer levels of w number at most four, and that an explicit rational instance attains four, so L*_Z=4. The proof proceeds by a strict diameter estimate bounding the real body to five levels, a five-level trichotomy reducing to three blocks, a nearest-integer inequality excluding the central block, and layer reversal and cell localization reducing the side blocks to eight modes; four easy modes are handled by a shared-cap inequality and the four hard modes by an exact Bernstein certificate with 17,640 positive rational coefficients. An appendix provides the rational attainment data and the claimed certificate minima.

Significance. If the certificate is independently confirmed, the result is a sharp, nontrivial constant in a fixed-direction arithmetic-width setting, and it usefully distinguishes the integer-occupied layer count from the real layer-intersection count. The analytic parts of the proof are self-contained and carefully argued, and the sharpness construction is explicit and checkable. The principal deficit is reproducibility: the certificate that excludes the four hard modes is not actually included in the manuscript, only per-mode minima and a hash digest, so the central theorem is currently conditional on an unverified finite computation. Once the certificate is supplied and the exact arithmetic is checked, the result merits publication.

major comments (3)

[Appendix A.2 and Proposition 5.2] The proof of the hard-mode inequality (32) depends on the assertion that all 17,640 Bernstein coefficients of the polynomial B are positive, with minimum 625/2048, but the manuscript prints neither the explicit polynomial B for each mode, nor the coefficient list, nor the program that produced the certificate. The per-mode minima and the hash digest are not independently checkable by a reader. Since Proposition 6.1 uses (32) in the final contradiction and the four hard modes have no alternative elementary bound, Theorem 7.1 cannot be considered established unless the exact program or a machine-readable certificate, including the eight polynomials and their Bernstein expansions, is provided and the digest can be reproduced.
[Section 5, proof of 7p-30g > 0] The displayed lower bound is incorrect as written: the bracket '725/4' should be '145/4', because with vartheta in [23/50,1/2] the minimum of 7(3-vartheta)^2 - 30 vartheta^2 is 145/4, not 725/4. The final value 85/8 follows only after this correction, and the derivation displayed in the paper does not justify the claimed inequality as printed. The same exact arithmetic should be re-checked throughout, since this line is used to establish the positivity of 7p-30g in the hard-mode argument.
[Section 8 and Appendix A.2] The reproducibility statement says that the proof relies on the rational coefficient inequalities and not on a digest or an unpublished software result, but the appendix does not actually exhibit the coefficients or the program. The statement overstates the state of reproducibility. The authors should either include the ancillary program exact_bernstein_certificate.py, or provide a complete machine-readable table of all 17,640 Bernstein coefficients (or the eight explicit polynomials together with the conversion code), so that a reader can verify the positivity claim independently.

minor comments (3)

[Section 5, notation] The symbol g is used both for the auxiliary function in Lemma 3.1 and for the quantity e^2 + r vartheta^2 in (27); renaming one of them would avoid confusion.
[Section 5, boundary conventions] The text says that boundary cases with e2=1/2 or mu=1/2 are retained only in a larger closed relaxation, but it would help to state explicitly which strict inequalities remain valid on that closed box and why the certificate's positivity on the larger box is sufficient for the strict inequality (32) on the original domain.
[Title and abstract] The title could state more prominently that the four-layer bound is established for the strict-failure branch of the specified deterministic reduction; the current wording is broader than the theorem.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the four-layer bound follows from fixed algorithmic reduction and explicit inequalities; the unprinted Bernstein certificate is a reproducibility caveat, not a circular step.

full rationale

No circularity found in the derivation chain. Theorem 7.1 is assembled from Proposition 3.2 (five-level trichotomy via the strict diameter estimate and beta > 4/3), Proposition 4.2 (nearest-integer inequality), Proposition 5.1 (side localization and the eight-mode list), Proposition 5.2 (Bernstein certificate for the hard modes), and Proposition 6.1 (shared-cap contradiction). Each of these steps uses inequalities derived from the defining setup: the deterministic delta = 3/4 reduction, the inner ellipse, the Babai point, and the strict-failure condition E_B > 1. No parameter is fitted to the target bound |J_Z(I)| <= 4. The functional w is selected by the reduction, and the occupied-level set J_Z(I) is defined from that fixed choice; the bound is then proved for that set. The attainment example in Appendix A.3 and A.4 is an explicit rational construction with exact arithmetic, not an optimized or fitted instance. The cited references are standard external works; there is no load-bearing self-citation chain and no uniqueness theorem imported from the authors. The manuscript is self-contained against external benchmarks, so the appropriate circularity score is 0. Two non-circular caveats are worth flagging under the review rules. First, Appendix A.2 records the Bernstein certificate only through per-mode minima and a digest, not the 17,640 coefficients or the ancillary program: 'For each of the eight modes it produces degree (4,6,8,6) and 2,205 coefficients. All coefficients are positive, and the smallest is 625/2048.' The text also says the digest 'is not used as a substitute for the rational coefficient checks themselves,' but the full coefficient list is not printed, so Proposition 5.2 is not independently checkable from the paper alone. This is a verification/correctness risk, not circularity. Second, the displayed lower bound '725/4' in Section 5 for 7(3-vartheta)^2 - 30 vartheta^2 appears to be an arithmetic typo for 145/4; the final value 85/8 is consistent with the corrected value, so the gap is not load-bearing. Section 9 explicitly limits the theorem's scope to the rational planar setting and the stated deterministic reduction; that is an honest scope restriction, not a circularity admission.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central derivation is self-contained: it starts from the defining quadratic form, slab, and affine coset, and the only non-standard inputs are the asserted properties of the deterministic delta = 3/4 reduction and the finite Bernstein certificate. There are no fitted constants; the attainment example is a construction. The certificate is an ad-hoc computational premise because the full output is not printed.

assumptions (4)

standard math A deterministic delta = 3/4 lattice reduction U in SL(2,Z) exists for the positive definite integral form G and yields -1/2 <= mu < 1/2, c_G >= 3a_G/4, and eta >= a_G/2.
Invoked at eq (11) to set alpha, beta, mu and the metric (13). Presented as a known exact reduction property, but stated without proof or citation.
domain assumption The instance lies in the nondegenerate strict-failure branch: R > 0, A < B, and alpha e1^2 + beta e2^2 > 1.
The theorem applies only to this branch (eq (15)); it excludes empty, singleton, and fixed-column cases and the case where the Babai point lies inside the inner ellipse.
domain assumption Tie-breaking convention: nearest-integer rounding sends every half-integer tie to the larger integer.
Fixes the deterministic Babai point and the half-open intervals for e1 and e2 in Section 2. Needed for the layer-reversal and closed-box arguments.
ad hoc to paper All 17,640 Bernstein coefficients of B for the four hard modes are positive, with minimum 625/2048.
Used in Proposition 5.2 to prove inequality (32). The exact rational computation is described but the full coefficient table is not printed in the manuscript, so the proof rests on the correctness of that computation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Sharp Four-Layer Theorem for Integer-Occupied Slices of a Planar Disk-Slab." pith.science (2026). https://pith.science/paper/A7PPH6RT

@misc{pith2026260812204,
  author       = {Pith},
  title        = {Pith review of: A Sharp Four-Layer Theorem for Integer-Occupied Slices of a Planar Disk-Slab},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/A7PPH6RT}},
  note         = {Machine review of arXiv:2608.12204}
}

read the original abstract

Let a rational positive-definite quadratic sublevel set on a full-rank affine lattice coset be intersected with a closed slab. A deterministic two-dimensional delta=3/4 reduction selects a primitive-dual integer functional and an inner ellipse. In the branch where the corresponding Babai point lies outside that ellipse, the lattice points in the disk-slab occupy at most four integer levels of the selected functional. An explicit rational instance attains four, so the bound is sharp. A strict diameter estimate reduces any counterexample to three blocks of five consecutive levels, and a nearest-integer inequality excludes the central block. Symmetry and cell localization reduce the two side blocks to eight modes: four follow from a shared-cap inequality, and four from an exact four-variable Bernstein certificate. The certificate contains 17,640 positive coefficients, with minimum 625/2048.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

11 extracted references · 10 canonical work pages

[1]

Closest point search in lattices.IEEE Transactions on Information Theory, 48(8):2201–2214, 2002

Erik Agrell, Thomas Eriksson, Alexander Vardy, and Kenneth Zeger. Closest point search in lattices.IEEE Transactions on Information Theory, 48(8):2201–2214, 2002

work page 2002
[2]

Exact Flatness Constant for One-Point Convex Bodies and the Discrete Isominwidth Problem: The Planar Case

Gennadiy Averkov, Giulia Codenotti, Ansgar Freyer, and Kyle Huang. Exact flatness constant for one-point convex bodies and the discrete isominwidth problem: The planar case, 2026. arXiv:2604.27260v2

work page Pith review arXiv 2026
[3]

On Lov´ asz’ lattice reduction and the nearest lattice point problem.Combina- torica, 6(1):1–13, 1986

L´ aszl´ o Babai. On Lov´ asz’ lattice reduction and the nearest lattice point problem.Combina- torica, 6(1):1–13, 1986

work page 1986
[4]

Generalised flatness constants: A framework applied in dimension 2, 2021

Gabriele Codenotti, Thomas Hall, and Johannes Hofscheier. Generalised flatness constants: A framework applied in dimension 2, 2021. arXiv:2110.02770

arXiv 2021
[5]

An arithmetic measure of width for convex bodies

Jes´ us A. De Loera, Brittney Marsters, and Christopher O’Neill. An arithmetic measure of width for convex bodies.arXiv preprint arXiv:2509.04726, 2025

work page Pith review arXiv 2025
[6]

Lattice-free polytopes and their diameter.Discrete & Compu- tational Geometry, 13(1):59–75, 1995

Michel Deza and Shmuel Onn. Lattice-free polytopes and their diameter.Discrete & Compu- tational Geometry, 13(1):59–75, 1995

work page 1995
[7]

Improved methods for calculating vectors of short length in a lattice, including a complexity analysis.Mathematics of Computation, 44:463–471, 1985

Ulrich Fincke and Michael Pohst. Improved methods for calculating vectors of short length in a lattice, including a complexity analysis.Mathematics of Computation, 44:463–471, 1985

work page 1985
[8]

C. A. J. Hurkens. Blowing up convex sets in the plane.Linear Algebra and its Applications, 134:121–128, 1990

work page 1990

Show all 11 references

[9]

Covering minima and lattice-point-free convex bodies.Annals of Mathematics, 128(3):577–602, 1988

Ravi Kannan and L´ aszl´ o Lov´ asz. Covering minima and lattice-point-free convex bodies.Annals of Mathematics, 128(3):577–602, 1988

1988
[10]

Lenstra, Jr

Hendrik W. Lenstra, Jr. Integer programming with a fixed number of variables.Mathematics of Operations Research, 8(4):538–548, 1983

1983
[11]

Lattice basis reduction: Improved practical algo- rithms and solving subset sum problems.Mathematical Programming, 66:181–199, 1994

Claus-Peter Schnorr and Martin Euchner. Lattice basis reduction: Improved practical algo- rithms and solving subset sum problems.Mathematical Programming, 66:181–199, 1994. 18

1994

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

Closest point search in lattices.IEEE Transactions on Information Theory, 48(8):2201–2214, 2002

Erik Agrell, Thomas Eriksson, Alexander Vardy, and Kenneth Zeger. Closest point search in lattices.IEEE Transactions on Information Theory, 48(8):2201–2214, 2002

work page 2002

[2] [2]

Exact Flatness Constant for One-Point Convex Bodies and the Discrete Isominwidth Problem: The Planar Case

Gennadiy Averkov, Giulia Codenotti, Ansgar Freyer, and Kyle Huang. Exact flatness constant for one-point convex bodies and the discrete isominwidth problem: The planar case, 2026. arXiv:2604.27260v2

work page Pith review arXiv 2026

[3] [3]

On Lov´ asz’ lattice reduction and the nearest lattice point problem.Combina- torica, 6(1):1–13, 1986

L´ aszl´ o Babai. On Lov´ asz’ lattice reduction and the nearest lattice point problem.Combina- torica, 6(1):1–13, 1986

work page 1986

[4] [4]

Generalised flatness constants: A framework applied in dimension 2, 2021

Gabriele Codenotti, Thomas Hall, and Johannes Hofscheier. Generalised flatness constants: A framework applied in dimension 2, 2021. arXiv:2110.02770

arXiv 2021

[5] [5]

An arithmetic measure of width for convex bodies

Jes´ us A. De Loera, Brittney Marsters, and Christopher O’Neill. An arithmetic measure of width for convex bodies.arXiv preprint arXiv:2509.04726, 2025

work page Pith review arXiv 2025

[6] [6]

Lattice-free polytopes and their diameter.Discrete & Compu- tational Geometry, 13(1):59–75, 1995

Michel Deza and Shmuel Onn. Lattice-free polytopes and their diameter.Discrete & Compu- tational Geometry, 13(1):59–75, 1995

work page 1995

[7] [7]

Improved methods for calculating vectors of short length in a lattice, including a complexity analysis.Mathematics of Computation, 44:463–471, 1985

Ulrich Fincke and Michael Pohst. Improved methods for calculating vectors of short length in a lattice, including a complexity analysis.Mathematics of Computation, 44:463–471, 1985

work page 1985

[8] [8]

C. A. J. Hurkens. Blowing up convex sets in the plane.Linear Algebra and its Applications, 134:121–128, 1990

work page 1990