Pith.

Dynamical Systems · arXiv:2611.17189 v1 · submitted 12 Nov 2026 · 8 references

Dimension drop for superlevel sets of the uniform Littlewood constant

Shahriar Mirzadeh

The paper says

“For (x,y) ∈ ℝ2 and Q ≥ 1, set $F_Q(x,y)=Q\min_{1\le q\le Q}\lVert qx\rVert\lVert qy\rVert$, where q ranges over positive integers and $\lVert\cdot\rVert$ denotes distance to the nearest integer. The set on which the uniform Littlewood constant $L=\limsup_{Q\to\infty}F_Q$ is positive has full Hausdorff dimension.”

more of the abstract

We prove that for every fixed c > 0 the superlevel set {L ≥ c} has Hausdorff dimension strictly less than two. More precisely, the Lebesgue measure of {(x,y) ∈ [0,1]2:FQ(x,y) ≥ c} is Oc(Q−κ(c)) for some κ(c) > 0 and all Q ≥ 2. The proof combines large deviations for one diagonal action on the space of unimodular lattices with effective equidistribution of expanding horospheres for another.

from the abstract

Pith says

Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded · not checked by a person

It can be wrong. Challenge this review or reply to it.

Holds up, with conditions.

4 major objections · 9 minor

How often objections from reviews like this one are right: not yet measured

Are you an author? What you can do.

Not checked: figures, code, what the 8 references say.

The paper says, at Section 2, equation (2.3)

The review did not quote this passage.

Pith says: Objection 1 of 4 · major

The large-deviation estimate (2.3) is cited from the unpublished arXiv preprint [2] (Kim-Mor, 2026). The paper states that [2, Theorem 1.2] applies because G is connected, semisimple, has finite center and no compact factors, and Γ is irreducible. However, the hypotheses of [2, Theorem 1.2] may require additional conditions (e.g., that b acts as a partially hyperbolic element with a spectral gap, or that the observable φ is sufficiently regular).

more of objection 1

The paper does not verify that these conditions are met for the specific element b = diag(e, e^{-1}, 1) and for the class of observables φ used in the proof. Since the entire quantitative estimate (Theorem 1.2) depends on this estimate, the reliance on an unpublished and potentially incomplete reference is a significant concern. The author should either provide a self-contained proof of the needed large-deviation bound, or confirm that [2] has been accepted for publication and that the hypotheses are satisfied. In the meantime, the result should be stated as conditional on the validity of [2].

The paper says, at Section 4, Theorem 4.1

The review did not quote this passage.

Pith says: Objection 2 of 4 · major

The proof of Theorem 4.1 uses the effective equidistribution estimate (E) with a test function f_N whose Sobolev norm grows like e^{BN}. The constant B depends on b, ℓ, and the choice of norms, but the estimate (E) requires that S_ℓ(f) be finite; it does not require uniformity in the support of f.

more of objection 2

However, the constant C_{ρ,x} in (E) is stated to be independent of f, but it may depend on the support of f through the requirement that T ≥ T_{ρ,x}. The paper asserts that T_{ρ,x} is independent of f, but the proof of Lemma 2.1 constructs T_{ρ,x} by taking the maximum of finitely many lower bounds from (2.5), each of which depends on the injectivity radius of the point z_i = u_{w_i} z. Since the points z_i depend on the support of ρ but not on f, this is fine. However, the application in Theorem 4.1 uses a fixed ρ (independent of N) and a fixed x, so T_{ρ,x} is fixed. The condition T ≥ T_{ρ,x} is then satisfied for all sufficiently large T, and the estimate (4.14) is valid. This point is handled correctly, but it would be helpful to state explicitly that T_{ρ,x} does not depend on N.

The paper says, at Section 5, Lemma 5.2

The review did not quote this passage.

Pith says: Objection 3 of 4 · major

The conversion from the measure estimate (5.2) to the covering estimate (5.3) uses the inequality F_Q(w') ≥ a/2 for w' in a square of side h = 1/M that meets A_Q(a). The Lipschitz constant in Lemma 5.1 is Q^2, so the condition Q^2 h ≤ a/4 ensures F_Q(w') ≥ a - a/4 = 3a/4, not a/2.

more of objection 3

The paper writes a - a/4 = a/2, which is correct (a - a/4 = 3a/4, but the text says a - a/4 ≥ a/2, which is true but the bound a/2 is weaker than 3a/4). The choice M = ⌈4Q^2/a⌉ gives h ≤ a/(4Q^2), so Q^2 h ≤ a/4, and then F_Q(w') ≥ a - a/4 = 3a/4 ≥ a/2. This is fine. However, the paper then claims that the whole square lies in A_Q(a/2). This is correct because F_Q(w') ≥ a/2. The subsequent step uses Leb(A_Q(a/2)) to bound N_Q(a) h^2. This is valid. The derivation of (5.3) is sound.

The paper says, at Section 5, equation (5.4)

The review did not quote this passage.

Pith says: Objection 4 of 4 · major

The inclusion eE_c ⊂ limsup A_{2^n}(c/4) is justified by the argument that if L(w) ≥ c, then there are arbitrarily large Q with F_Q(w) ≥ c/2, and then F_{2^n}(w) ≥ c/4 for the corresponding n. This uses the fact that the set of denominators up to 2^n is contained in the set up to Q, and that F_{2^n}(w) = 2^n min_{q ≤ 2^n} ... ≥ (2^n/Q) F_Q(w).

more of objection 4

Since Q < 2^{n+1}, we have 2^n/Q > 1/2, so F_{2^n}(w) > (1/2)(c/2) = c/4. This is correct. The limsup inclusion is then standard.

9 minor objections

The paper says, at Abstract and Introduction

Pith says: minor objection 1 of 9

The abstract states that the proof combines large deviations for one diagonal action with effective equidistribution of expanding horospheres for another. This is accurate but could be slightly more explicit about which actions are used (b_s and a_T).

The paper says, at Section 2, equation (2.3)

Pith says: minor objection 2 of 9

The large-deviation estimate is stated for a fixed φ ∈ C_c^∞(X) with a constant c_φ that may depend on φ. The paper later applies it to a fixed φ (the cutoff for Ω_c), so this is consistent. However, the statement 'for every real-valued φ ∈ C_c^∞(X) there is c_φ > 0 such that ...' is ambiguous about whether c_φ depends on φ only or also on ε. The reference [2] should clarify this.

show 7 more

The paper says, at Section 3, Lemma 3.1

Pith says: minor objection 3 of 9

The proof that q ≠ 0 uses the fact that |p|,|r| < √c < 1, which forces p = r = 0, contradicting (p,r,q) ≠ 0. This is correct. The argument that 1 ≤ |q| < Q uses the third coordinate bound |e^{-T} q| < 1, so |q| < e^T = Q, and q ≠ 0 implies |q| ≥ 1. This is sound.

The paper says, at Section 4, equation (4.9)

Pith says: minor objection 4 of 9

The constant β is defined by ∥Ad(b^j)∥ ≤ e^{β j}. Since b = diag(e, e^{-1}, 1), the operator norm of Ad(b^j) on the Lie algebra sl_3(R) grows like e^{j} (the largest eigenvalue of Ad(b) is e). So β can be taken as 1. The paper uses β generically, which is fine.

The paper says, at Section 4, equation (4.13)

Pith says: minor objection 5 of 9

The bound S_ℓ(f_N) ≤ C_{φ,ε} e^{BN} is derived from the chain rule and the estimate (4.12). The constant C_{φ,ε} may depend on φ and ε but not on N. This is acceptable.

The paper says, at Section 4, Corollary 4.2

Pith says: minor objection 6 of 9

The choice θ = m/4 is used to ensure that the low-return event is contained in the deviation event with ε = m/2. This is valid because if R_N ≤ θN, then P_N^φ ≤ m/4, so |P_N^φ - m| ≥ 3m/4 > m/2. The corollary then follows from Theorem 4.1.

The paper says, at Section 5, Lemma 5.2

Pith says: minor objection 7 of 9

The final step that dim_H E_c = dim_H eE_c uses the fact that the quotient map R^2 → T^2 is locally bi-Lipschitz and that Hausdorff dimension is preserved under bi-Lipschitz maps and countable unions. This is standard and correctly stated.

The paper says, at General

Pith says: minor objection 8 of 9

The paper uses the notation A_Q(c) for the finite-scale set and later uses A_Q(a/2) etc. This is consistent. The proof of Theorem 1.2 at the end of Section 4 uses Corollary 4.2 with Ω = Ω_c and K = {u_w : w ∈ [0,1]^2}. The identification of Lebesgue measure on T^2 with the Haar measure on U restricted to [0,1]^2 is valid up to a null set (the boundary). This is noted.

The paper says, at References

Pith says: minor objection 9 of 9

Reference [2] is an unpublished arXiv preprint from 2026. The paper should note that the result is conditional on the correctness of [2]. Reference [5] is also an arXiv preprint from 2026. These are acceptable for a preprint, but for journal publication the author should ensure that all cited results are either published or accepted.

What would settle it

Independently verify the large-deviation estimate (2.3) for the specific function φ used in Corollary 4.2 (a smooth cutoff supported in Ω_c) by either (a) deriving it from the spectral gap of the b-action on SL_3(R)/SL_3(Z) using known results (e.g., exponential mixing of the geodesic flow on SL_3(R)/SL_3(Z) and the method of [2]), or (b) confirming that the preprint [2] has been accepted for publication and the estimate holds as stated. If the estimate cannot be verified, the proof of Theorem 1.2 collapses.

from the skeptic's reading

How the review went

3 readings by deepseek-v4-flash · 13 Nov 2026

  1. Reader holds up, with conditions confidence moderate
  2. Skeptic partly agrees 2 min 24 s · 13,121 tokens read, 1,258 written
  3. Referee major revision: 4 major objections, 9 minor 17 s · 13,486 tokens read, 2,397 written

Not run on this paper: the editor, the simulated authors' defence.

Pith did not check this review's objections against the paper.

The review in full

The strongest objection

The central quantitative estimate (Theorem 1.2) depends on the exponential large-deviation bound (2.3), which is cited from the unpublished arXiv preprint [2] (Kim–Mor, 2026). The proof of Theorem 4.1 requires this bound to hold uniformly for observables φ whose Sobolev norms grow like e^{c φ j} (see (4.13) and the chain-rule estimate leading to it). The function φ used in Corollary 4.2 is a smooth cutoff supported in the open set Ω_c; its derivatives are bounded, but the Sobolev norm of the derived function f_N grows exponentially with N because the support of f_N expands under the b-action. The large-deviation estimate (2.3) as stated in the paper applies to a fixed φ ∈ C_c^∞(X) with a constant c_φ that may depend on φ. However, the argument in Theorem 4.1 applies (LD) to the same fixed φ (the original cutoff), not to f_N; the large-deviation estimate is used only for φ, not for f_N. The Sobolev-norm growth (4.13) is used only in the equidistribution step (E), not in the large-deviation step. Therefore the concern about uniformity of (2.3) for observables with growing Sobolev norms is misplaced: the large-deviation estimate is applied to the fixed φ, and the Sobolev growth affects only the equidistribution estimate, which is already uniform in the test function (see (E) and Lemma 2.1). The paper's own text confirms this: 'The large-deviation estimate of Kim and Mor [2], applied to b = b_1, bounds the Haar measure of points whose b-orbits visit Ω_c too infrequently.' The function φ is fixed once Ω_c is chosen. Thus the reliance on [2] is for a standard large-deviation bound for a fixed smooth compactly supported function, which is a plausible result. The paper acknowledges that [2] is unpublished, but this is a common practice in mathematics; the result is cited, not proved.

The reader's weakest_assumption identifies the reliance on the unpublished preprint [2] as the main concern. I agree that this is a potential vulnerability, but I find that the paper's argument does not require the large-deviation estimate to hold for observables with growing Sobolev norms, contrary to the reader's suggestion. The large-deviation estimate is applied to a fixed smooth compactly supported function φ, and the Sobolev-norm growth in (4.13) is used only in the equidistribution step, which is already uniform. Therefore the concern is less severe than the reader implies. The paper's own structure is sound, and the conditional verdict is appropriate. I do not see a need to change the verdict. The paper's limitations are honestly stated. The concrete test I propose is to verify the large-deviation estimate for the specific φ, which is the only external input that is not yet peer-reviewed. If that estimate holds, the paper's argument is complete. The paper's own Lemma 3.1, Lemma 5.1, Lemma 5.2, and the transfer theorem (Theorem 4.1) are all verified within the paper. The effective equidistribution (2.4) is from a published source [3]. The paper's central claim is new and does not reduce to the cited inputs. The paper's own sentences (b21, b2, b3) state the main results clearly. The paper's own limitations (b29, b156) are acknowledged. I therefore recommend no change to the verdict.

The referee's report

If correct, the result resolves a natural question left open by the recent constructions of counterexamples to the uniform Littlewood conjecture: although the set of counterexamples has full Hausdorff dimension, no fixed positive superlevel set does. This is a nontrivial application of homogeneous dynamics to Diophantine approximation, and the transfer theorem (Theorem 4.1) may be of independent interest. The paper is clearly written and the argument is carefully organized. The main strength is the clean combination of two different dynamical estimates (large deviations and effective equidistribution) to obtain a dimension drop.

Pith's reading of the paper (a machine's summary, not the authors')

The classical Littlewood conjecture in number theory asks whether every pair (x,y) of real numbers admits integers q for which q times the product of the distances from qx and qy to the nearest integer tends to zero. The uniform Littlewood problem imposes a stronger requirement: at every large scale Q one should find some q at most Q with that product smaller than epsilon/Q. That uniform statement is false. The set E of pairs where the limit-superior constant L(x,y) is positive is not empty, and recent work shows E has full Hausdorff dimension, meaning it is as big as the whole plane in a fractal sense.

This paper proves the full dimension cannot be witnessed at any fixed level. For each fixed positive threshold c, the set E_c = {L >= c} has Hausdorff dimension strictly below two. The engine is a counting estimate: the measure of pairs where the finite-scale quantity F_Q(x,y) stays above c decays like a power of Q, namely O(Q^{-κ(c)}). A dyadic covering argument converts that decay into the dimension bound.

The proof works on the space of unimodular lattices in three dimensions. A Diophantine condition on (x,y) is encoded by whether a certain diagonal flow keeps the lattice away from a fixed set of lattices having a short vector. A large-deviations estimate controls how rarely orbits visit that special set, and an effective equidistribution result transfers this control from all lattices to the two-dimensional parameter family (x,y). A smooth majorant with controlled growth and a dyadic sum finish the proof. The method gives no explicit value for κ(c).

Core claim. Theorem 1.1: for every c > 0 there exists δ(c) > 0 such that dim_H E_c ≤ 2 - δ(c) < 2, where E_c = {(x,y) in R^2 : L(x,y) >= c} and L is the uniform Littlewood constant. If correct, each fixed positive superlevel set of L has Hausdorff dimension strictly below two, so the full-dimensional counterexample set E = union of E_{1/n} attains dimension 2 only in the limit c down to 0. The supporting finite-scale statement, Theorem 1.2, gives Leb(A_Q(c)) ≤ C_c Q^{-κ(c)} for some κ(c) in (0,1], which drives the dimension bound 2 - κ(c/8)/2.

The premise it rests on. The quantitative chain hinges on the exponential large-deviation estimate (2.3), hypothesis (LD) of Theorem 4.1, imported from the unpublished v1 arXiv preprint [2] (Kim-Mor, 2026). The proof requires the bound to hold uniformly for observables whose Sobolev norms grow like e^{cφ j}, combined with the Sobolev-growth bound (4.13), so that both terms in (4.14) decay with T. A failure of uniformity in [2], or a correction or withdrawal of [2], would break Theorem 4.1 and with it Theorem 1.2 and Theorem 1.1. Location: Section 2 equation (2.3); Theorem 4.1 hypotheses (LD) and equation (4.6); reference [2]. Secondary premise: the effective equidistribution (2.4) from [3] used in Lemma 2.1.

Check this review

Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.

How often reviews like this one are right

Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.

How often objections from reviews like this one are right: not yet measured

Every value above, in the review's public record (JSON)

Reviews by people

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review

How Pith checks papers · The full audit of this paper

The paper

Abstract

For (x,y) ∈ ℝ2 and Q ≥ 1, set $F_Q(x,y)=Q\min_{1\le q\le Q}\lVert qx\rVert\lVert qy\rVert$, where q ranges over positive integers and $\lVert\cdot\rVert$ denotes distance to the nearest integer. The set on which the uniform Littlewood constant $L=\limsup_{Q\to\infty}F_Q$ is positive has full Hausdorff dimension. We prove that for every fixed c > 0 the superlevel set {L ≥ c} has Hausdorff dimension strictly less than two. More precisely, the Lebesgue measure of {(x,y) ∈ [0,1]2:FQ(x,y) ≥ c} is Oc(Q−κ(c)) for some κ(c) > 0 and all Q ≥ 2. The proof combines large deviations for one diagonal action on the space of unimodular lattices with effective equidistribution of expanding horospheres for another.

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.

abstract compiled · 2026-11-14

The authors' main claim, as the classifier picked it out 1.00

“We prove that for every fixed $c>0$ the superlevel set $\{L\ge c\}$ has Hausdorff dimension strictly less than two.”

How the main claim is stated: asserted 1.00

What the abstract itself states, as probabilities: claims first 0.04 · states limitation 0.44 · human participants 0.01 · controlled comparison 0.02 · randomized assignment 0.01 · states quantitative result 0.93 · compares with prior results 0.38 · mentions code or data availability 0.01

The abstract, sentence by sentence

This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.

SentenceRoleThe authors' own workStates the main conclusionStrength
For $(x,y)\in\mathbb{R}^2$ and $Q\ge 1$, set $F_Q(x,y)=Q\min_{1\le q\le Q}\lVert qx\rVert\lVert qy\rVert$, where $q$ ranges over positive integers and $\lVert\cdot\rVert$ denotes distance to the nearest integer. background 0.99 0.18 0.02 asserted 0.56
The set on which the uniform Littlewood constant $L=\limsup_{Q\to\infty}F_Q$ is positive has full Hausdorff dimension. background 0.79 0.24 0.11 asserted 1.00
We prove that for every fixed $c>0$ the superlevel set $\{L\ge c\}$ has Hausdorff dimension strictly less than two. result 0.94 0.95 0.87 asserted 1.00
More precisely, the Lebesgue measure of $\{(x,y)\in[0,1]^2:F_Q(x,y)\ge c\}$ is $O_c(Q^{-\kappa(c)})$ for some $\kappa(c)>0$ and all $Q\ge 2$. result 1.00 0.90 0.14 asserted 1.00
The proof combines large deviations for one diagonal action on the space of unimodular lattices with effective equidistribution of expanding horospheres for another. method 1.00 0.75 0.03 asserted 0.98

What kind of paper the classifier reports this is: theoretical result 1.00 · evidence from proof or derivation 1.00

What was examined, and what was not

Sources read

  • arXiv listing abstract v1

Not examined

  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers

Levels run

  • P abstract

Not run

  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

How the labels were made

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-14: 1 request, 4,541 input tokens.

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON

References (8)
  1. [2] T. Kim and R. Mor,Large deviations and applications in homogeneous dynamics, arXiv:2610.32814v1 (2026) on PithPith's reviewarXiv
  2. [3] D. Kleinbock and S. Mirzadeh,Dimension drop for diagonalizable flows on homogeneous spaces, J. Mod. Dyn.20(2024), 441–478, doi:10.3934/jmd.2024012 on Pithdoi:10.3934/jmd.2024012
  3. [1] P. Bandi, R. Fregoli and D. Kleinbock,Submanifold-genericity ofR d-actions and uniform multiplicative Diophantine approximation, arXiv:2504.02258 (2025) on PithPith's reviewarXiv
  4. [4] D. Kleinbock and S. Mirzadeh,Dimension bounds for escape on average in homogeneous spaces, Discrete Contin. Dyn. Syst.45(2025), no. 5, 1653–1671, doi:10.3934/dcds.2024141 on Pithdoi:10.3934/dcds.2024141
  5. [5] N. Moshchevitin,A note on uniform version of Littlewood inequality and Fibonacci numbers, arXiv:2605.26188 (2026) on PithPith's reviewarXiv
  6. [6] V. Neckrasov, C. Wu and B. Yang,Winning property of counterexamples to uniform Little- wood’s conjecture, arXiv:2608.24401v2 (2026) on PithPith's reviewarXiv
  7. [7] J. Schleischitz,Disproof of the uniform Littlewood conjecture, arXiv:2603.12611 (2026) on PitharXiv
  8. [8] N. Shulga,The uniform Littlewood conjecture fails on a set of positive Hausdorff dimension, arXiv:2608.25059 (2026). Department of Mathematics, Brandeis University, 415 South Street, W altham, MA 02453, USA Email address:shahmir@brandeis.edu on PithPith's reviewarXiv

Challenges and discussion

Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.

No comments yet.

Sign in to comment