REVIEW 2 major objections 3 minor 30 references
Semi-algebraic discrepancy estimates for multi-frequency shift sequences with applications to quantum dynamics
T0 review · 2 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read The paper proves near-optimal bounds on how often multi-frequency shift sequences hit curved sets, and uses them to control wave-packet spreading in long-range quasi-periodic Schrödinger operators.
desk verdict The upper-bound discrepancy estimate and quantum-dynamics corollary are clean and correct; Theorem 1.3's lower-bound sharpness proof miscounts and fails as written. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The mechanism that carries the upper bound is the one-dimensional weak Diophantine condition $\|n\alpha\|_{\mathbb{T}^b}\ge\gamma/|n|^\tau$ for all $n\in\mathbb{Z}\setminus\{0\}$. It makes the covering argument local: any two distinct points $\theta+n\alpha$ and $\theta+n'\alpha$ lying in the same small ball would force $\|(n-n')\alpha\|\le2\epsilon$, contradicting the condition once $\epsilon=\gamma/(2N^\tau)$. Combined with the covering lemma for semi-algebraic sets (a set of degree $B$ and measure at most $\epsilon^b$ is covered by $B C(b)\epsilon^{1-b}$ balls), this gives at most one counted index per ball and hence the bound $B C(b)N^{\tau(b-1)}$. The lower bound instead constructs, via a metric discrepancy lemma for Kronecker sequences, $b-1$ integer multiples $n_i\alpha$ with $\|\{n_i\alpha\}\|_\infty\le N^{-r/b}$ that are linearly independent and span a hyperplane $S$; integer combinations of these vectors with $1\le k_i n_i\le N^b$ are then counted as points of $S$. For quantum dynamics, the load-bearing transfer is the large deviation theorem for Green's functions of long-range quasi-periodic operators (which supplies a semi-algebraic exceptional set $\Theta_{N_1}$ of measure at most $e^{-N^c}$) together with a criterion that bounds the $p$-th moment by $(\log T)^{p/\delta}$ whenever the number of bad sites in an interval of length $N$ is at most $N^{1-\delta}$; the discrepancy estimate supplies $\delta=1-\tau(b-1)-\varepsilon$.
What would settle it
Verify the Section 3.2 counting identity in dimension $b=2$: Proposition 3.2 produces $n_1\in[1,N^{2(1+2\varepsilon)/3}]$ with $\|\{n_1\alpha\}\|_\infty\le N^{-1/3}$. The proof counts every $k_1$ satisfying $1\le k_1 n_1\le N^2$ and $k_1\|\{n_1\alpha\}\|_\infty<1/2$, and then treats $k_1 n_1$ as lying in $[1,N]$. Computing, for an explicit Diophantine $\alpha$ and large $N$, whether the number of such $k_1$ with $1\le k_1 n_1\le N$ is at least $N^{1/3-\varepsilon}$ would settle Theorem 1.3; a shortfall means the product count overcounts points outside the window.
Extended reading notes
Core claim
The paper's central claim is that the almost-everywhere semi-algebraic discrepancy exponent for multi-frequency shift sequences is $1-1/b$, with a lower-bound exponent $1-1/(b+1)$ that approaches it as $b\to\infty$. Theorem 1.2 says that for $\alpha\in WDC(\gamma,\tau)$, every semi-algebraic $S\subseteq[0,1]^b$ of degree $B$ and Lebesgue measure at most $\eta$, and every $N$ with $\log N < (2\tau b)^{-1}\log(1/\eta)$, satisfies $\#\{1\le n\le N:\theta+n\alpha\bmod\mathbb{Z}^b\in S\}\lesssim B C(b) N^{\tau(b-1)}$; taking $\tau=1/b+\varepsilon$ gives $\le N^{(b-1)/b+\varepsilon}$ for almost every $\alpha$. Theorem 1.3 complements this with a hyperplane $S\subseteq[0,1]^b$ for which $\#\{1\le n\le N:n\alpha\bmod\mathbb{Z}^b\in S\}\ge N^{(b-1)/(b+1)-\varepsilon}$ for almost every $\alpha$ and all sufficiently large $N$. The transfer to quantum dynamics is Theorem 1.5: for $\alpha\in WDC(\tau)\cap DC(\tau')$ and nonconstant real-analytic $V$, there is a threshold $\lambda_0$ such that for $\lambda>\lambda_0$, $\sup_{\theta}\langle |X_{H_{\theta,\alpha}}|^p\rangle_\psi(T)\le(\log T)^{p/(1-\tau(b-1))+\varepsilon}$, and Corollary 1.6 gives the almost-everywhere version $(\log T)^{pb+\varepsilon}$.
Load-bearing premise
The lower-bound proof assumes that every integer combination $n=\sum_{i=1}^{b-1}k_i n_i$ with $1\le k_i n_i\le N^b$ and $k_i\|\{n_i\alpha\}\|_\infty<1/b$ automatically lies in the counting window $1\le n\le N$; the constraints only force $n_i\le N^{1+2\varepsilon}$ and $k_i\le N^b/n_i$, so $n$ can be as large as $(b-1)N^b$, and the proof drops the $n\le N$ restriction when passing from the counting set to the product of $k_i$-counts.
Editorial extensions
If this is right
- For almost every $\alpha\in\mathbb{T}^b$, every semi-algebraic set of small measure is visited at most $N^{(b-1)/b+\varepsilon}$ times among the first $N$ shifts, improving the previous best exponents $1-1/(b^2(b-1)+b)$ and $1-1/(2b)$ to $1-1/b$.
- For $\alpha\in WDC(\tau)\cap DC(\tau')$ and nonconstant real-analytic $V$, the phase-uniform moment satisfies $\sup_\theta\langle|X|^p\rangle_\psi(T)\le(\log T)^{p/(1-\tau(b-1))+\varepsilon}$; for almost every $\alpha$ this reads $(\log T)^{pb+\varepsilon}$.
- The lower-bound hyperplane construction shows the upper estimate cannot be improved by more than a factor of roughly $(b+1)/b$ in the exponent, so the discrepancy result is asymptotically sharp as $b\to\infty$.
- The counting argument gives a uniform-in-phase statement for all $\theta$, rather than only a phase-averaged localization bound.
- The proof of the upper bound works under the weaker one-dimensional weak Diophantine condition, which makes the argument shorter than earlier approaches based on the full multi-frequency Diophantine class.
Reading between the lines
- The gap between the upper exponent $(b-1)/b$ and the constructed lower exponent $(b-1)/(b+1)$ suggests the almost-everywhere discrepancy exponent may be strictly between these values; replacing hyperplanes by higher-degree semi-algebraic witnesses is a concrete way to test whether the upper bound is attainable.
- Because the covering argument uses only the one-dimensional small-multiples condition $\|n\alpha\|\ge\gamma/|n|^\tau$, the same upper bound should hold for Kronecker sequences under any nonlinear reparametrization that preserves this condition, provided the target sets remain semi-algebraic after the reparametrization.
- The logarithmic moment bound in Theorem 1.5 is an upper bound; if quasi-periodic long-range models in this regime instead exhibit power-law spreading, the discrepancy method alone cannot reveal it, since its input is already asymptotically sharp in dimension.
- A $C^k$ (finite-smoothness) version of the potential would remove the semi-algebraic approximation of the exceptional set on which the proof relies; whether the exponent persists for smooth but non-analytic potentials is an open question the paper does not address.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies semi-algebraic discrepancy for multi-frequency Kronecker sequences. Theorem 1.2 gives an upper bound for the number of visits of the orbit to a semi-algebraic set of small measure, under a weak Diophantine condition on the frequency. Theorem 1.3 claims a matching lower bound for almost every frequency, using a hyperplane constructed from b-1 small multiples of alpha. Theorem 1.5 applies the upper bound to long-range quasi-periodic Schr\"odinger operators, combining a large deviation theorem from [Liu22] with a dynamical reduction from [Liu23] to obtain logarithmic upper bounds on position moments. The upper-bound proof in Section 2 is a clean covering argument using Bourgain's semi-algebraic covering lemma; the quantum application is a direct combination of imported theorems. The lower-bound proof in Section 3.2 has a serious gap: it counts tuples of integers without enforcing that the resulting index n is at most N.
Significance. If the upper-bound theorem and the quantum application are correct, they improve known semi-algebraic discrepancy exponents from roughly N^{1-1/b^2+\varepsilon} to N^{1-1/b+\varepsilon} and give corresponding logarithmic quantum-dynamical bounds. The upper-bound proof is transparent and uses standard tools, and the quantum application is mechanical given the imported LDT and dynamical reduction; these parts are worth publishing if the issue with the lower bound is resolved. However, the advertised 'asymptotically sharp' aspect rests entirely on Theorem 1.3, whose proof is invalid as written, so the sharpness claim is not currently supported. The paper does not ship machine-checked proofs, but the main upper-bound derivation is simple enough to be checked by hand.
major comments (2)
- [§3.2, proof of Theorem 3.3] The step immediately after (19) counts all tuples (k_1,...,k_{b-1}) satisfying 1 \le k_i n_i \le N^b and k_i\|\{n_i\alpha\}\|_\infty < 1/b, and treats this count as a lower bound for #{1 \le n \le N : n\alpha \bmod \mathbb{Z}^b \in S}. This is only legitimate if the associated integer index n = \sum_i k_i n_i satisfies 1 \le n \le N. No such constraint is imposed: the written conditions give only n \le (b-1)N^b, and even reading the displayed bound as the intended N, one gets only n \le (b-1)N. Thus many counted tuples correspond to indices outside the window of the theorem, and the product lower bound is not a lower bound for the left-hand side. The proof needs a counting argument for the tuples that additionally satisfy \sum_i k_i n_i \le N; under the R-linear independence condition (18) this is a simplex count of order N^{b-1}/((b-1)!\prod_i n_i), which would yield the claimed exponent, but that argument is absent.
- [§4, proof of Theorem 1.5] The application of Theorem 4.2 requires an exponent \delta > 0 such that #B_{N,N^\varepsilon} \le N^{1-\delta}. From the displayed estimate (25) the proof obtains 1-\delta = \tau(b-1)+\varepsilon, so the proof requires \tau(b-1) < 1. Theorem 1.5 is stated for arbitrary \alpha \in WDC(\tau) \cap DC(\tau') with only \tau \ge 1/b; for \tau(b-1) \ge 1 the argument does not apply and the displayed exponent p/(1-\tau(b-1)) in the theorem is nonpositive. The statement should either add the hypothesis \tau < 1/(b-1) or be restricted to that range, which is sufficient for the full-measure Corollary 1.6.
minor comments (3)
- [Abstract and Corollary 1.6] The notation 'T^d' in Corollary 1.6 should be 'T^b' to match the dimension used in the rest of the paper.
- [§3.2, display after (19)] The constraint is written as 1 \le k_i n_i \le N^b, while the subsequent lower bound uses N^{1-r(1+2\varepsilon)}; these two expressions are inconsistent and need to be reconciled.
- [§3.1, Eq. (16)] The expression '|\langle w, \{n_k\alpha\}\rangle|^p' appears to contain a spurious power p; the surrounding inequalities suggest the absolute value should not be raised to a variable power.
Circularity Check
No significant circularity: the discrepancy estimates rest on independent external tools, and the quantum-dynamics application uses prior published theorems as black boxes.
full rationale
The main discrepancy upper bound (Theorem 1.2) is derived from Bourgain's semi-algebraic covering lemma (Lemma 2.1, cited externally) and the weak Diophantine separation property; no parameter is fitted to the target count and the conclusion does not appear as an input. The lower-bound construction (Theorem 1.3) uses Schmidt's discrepancy theorem as its external input and builds the hyperplane from independent lattice points; the counting step is a geometric product count, not a restatement of the desired bound. The quantum-dynamics section invokes Theorems 4.1 and 4.2 from [Liu22] and [Liu23] as published external theorems with stated assumptions (DC conditions and LDT hypotheses) that do not include the conclusion of Theorem 1.5; although these are self-citations, they are real evidence rather than circular support. The noted concern about the range constraint n = sum k_i n_i <= N in the proof of Theorem 1.3 is a potential correctness gap in the sharpness argument, not a circularity of the fitted-input or self-definitional type, and it does not affect the independent derivation of the upper-bound discrepancy estimate.
Assumptions & free parameters
assumptions (5)
- standard math Bourgain's semi-algebraic covering lemma (Corollary 9.6 of [Bou05])
- standard math Schmidt's discrepancy theorem for Kronecker sequences (Lemma 3.1, [Sch64])
- domain assumption Large Deviation Theorem for long-range quasi-periodic Schrödinger operators (Theorem 4.1, [Liu22])
- domain assumption Semi-algebraicity and degree bound for the LDT exceptional set (Bourgain p.56 argument)
- domain assumption Dynamical reduction from bad-site counting to moment bounds (Theorem 4.2, [Liu23, Cor. 2.3])
Cite this review
Pith. "Pith review of Semi-algebraic discrepancy estimates for multi-frequency shift sequences with applications to quantum dynamics." pith.science (2026). https://pith.science/paper/OKDR5B7N
@misc{pith2026250719783,
author = {Pith},
title = {Pith review of: Semi-algebraic discrepancy estimates for multi-frequency shift sequences with applications to quantum dynamics},
year = {2026},
howpublished = {\url{https://pith.science/paper/OKDR5B7N}},
note = {Machine review of arXiv:2507.19783}
}
read the original abstract
We establish asymptotically sharp semi-algebraic discrepancy estimates for multi-frequency shift sequences. As an application, we obtain novel upper bounds for the quantum dynamics of long-range quasi-periodic Schr\"odinger operators.
Reference graph
Works this paper leans on
-
[1]
J. Bourgain and M. Goldstein, On nonperturbative localization with quasi-periodic potential, Ann. of Math. (2) 152 (2000), no. 3, 835--879. 1815703
work page 2000
-
[2]
J. Bourgain and S. Jitomirskaya, Anderson localization for the band model, Geometric aspects of functional analysis, Lecture Notes in Math., vol. 1745, Springer, Berlin, 2000, pp. 67--79. 1796713
work page 2000
-
[3]
G. Binyamini and D. Novikov, Complex cellular structures, Ann. of Math. (2) 190 (2019), no. 1, 145--248. 3990603
work page 2019
-
[4]
J. Bourgain, Green's function estimates for lattice S chr\"odinger operators and applications , Annals of Mathematics Studies, vol. 158, Princeton University Press, Princeton, NJ, 2005. 2100420
work page 2005
-
[5]
A. Cai, H. Lv, Y. Shan, and X. Wang, Arithmetic (k-d/2) -polynomial dynamical localization for k power-law quasi-periodic long-range operators on Z ^d , preprint (2025)
work page 2025
-
[6]
D. Damanik, M. Lukic, and W. Yessen, Quantum dynamics of periodic and limit-periodic J acobi and block J acobi matrices with applications to some quantum many body problems , Comm. Math. Phys. 337 (2015), no. 3, 1535--1561. 3339185
work page 2015
-
[7]
D. Damanik and S. Tcheremchantsev, Upper bounds in quantum dynamics, J. Amer. Math. Soc. 20 (2007), no. 3, 799--827. 2291919
work page 2007
-
[8]
, Quantum dynamics via complex analysis methods: general upper bounds without time-averaging and tight lower bounds for the strongly coupled F ibonacci H amiltonian , J. Funct. Anal. 255 (2008), no. 10, 2872--2887. 2464194
work page 2008
Show all 30 references
-
[9]
Fillman, Ballistic transport for limit-periodic J acobi matrices with applications to quantum many-body problems , Comm
J. Fillman, Ballistic transport for limit-periodic J acobi matrices with applications to quantum many-body problems , Comm. Math. Phys. 350 (2017), no. 3, 1275--1297. 3607475
2017
-
[10]
Theory Adv
, Ballistic transport for periodic J acobi operators on Z ^d , From operator theory to orthogonal polynomials, combinatorics, and number theory---a volume in honor of L ance L ittlejohn's 70th birthday, Oper. Theory Adv. Appl., vol. 285, Birkh\"auser/Springer, Cham, [2021] 202...
2021
-
[11]
Ge and I
L. Ge and I. Kachkovskiy, Ballistic transport for one-dimensional quasiperiodic S chr\" o dinger operators , Comm. Pure Appl. Math. 76 (2023), no. 10, 2577--2612. 4630598
2023
-
[12]
Gromov, Entropy, homology and semialgebraic geometry, Ast\'erisque (1987), no
M. Gromov, Entropy, homology and semialgebraic geometry, Ast\'erisque (1987), no. 145-146, 5, 225--240, S\'eminaire Bourbaki, Vol.\ 1985/86. 880035
1987
-
[13]
L. Ge, J. You, and Q. Zhou, Exponential dynamical localization: criterion and applications, Ann. Sci. \'Ec. Norm. Sup\'er. (4) 56 (2023), no. 1, 91--126. 4637128
2023
-
[14]
Han and S
R. Han and S. Jitomirskaya, Quantum dynamical bounds for ergodic potentials with underlying dynamics of zero topological entropy, Anal. PDE 12 (2019), no. 4, 867--902. 3869380
2019
-
[15]
Jitomirskaya, H
S. Jitomirskaya, H. Kr\"uger, and W. Liu, Exact dynamical decay rate for the almost M athieu operator , Math. Res. Lett. 27 (2020), no. 3, 789--808. 4216568
2020
-
[16]
Jitomirskaya, W
S. Jitomirskaya, W. Liu, and L. Mi, Sharp palindromic criterion for semi-uniform dynamical localization, arXiv preprint arXiv:2410.21700 (2024)
2024 arXiv
-
[17]
Jitomirskaya and M
S. Jitomirskaya and M. Powell, Logarithmic quantum dynamical bounds for arithmetically defined ergodic S chr \"o dinger operators with smooth potentials , Analysis at Large: Dedicated to the Life and Work of Jean Bourgain, Springer, 2022, pp. 173--201
2022
-
[18]
Jitomirskaya and B
S. Jitomirskaya and B. Simon, Operators with singular continuous spectrum. III . A lmost periodic S chr\"odinger operators , Comm. Math. Phys. 165 (1994), no. 1, 201--205. 1298948
1994
-
[19]
Jitomirskaya and S
S. Jitomirskaya and S. Zhang, Quantitative continuity of singular continuous spectral measures and arithmetic criteria for quasiperiodic S chr\" o dinger operators , J. Eur. Math. Soc. (JEMS) 24 (2022), no. 5, 1723--1767. 4404788
2022
-
[20]
A. Ya. Khinchin, Continued fractions. ( U niversity of C hicago P ress) , The Mathematical Gazette 49 (1964), no. 369
1964
-
[21]
Liu, Quantitative inductive estimates for G reen's functions of non-self-adjoint matrices , Anal
W. Liu, Quantitative inductive estimates for G reen's functions of non-self-adjoint matrices , Anal. PDE 15 (2022), no. 8, 2061--2108. 4546503
2022
-
[22]
, Power law logarithmic bounds of moments for long range operators in arbitrary dimension, J. Math. Phys. 64 (2023), no. 3, Paper No. 033508, 11. 4564259
2023
-
[23]
H. L. Montgomery, Ten lectures on the interface between analytic number theory and harmonic analysis, CBMS Regional Conference Series in Mathematics, vol. 84, Published for the Conference Board of the Mathematical Sciences, Washington, DC; by the American Mathematical Society,...
1994
-
[24]
W. M. Schmidt, Metrical theorems on fractional parts of sequences, Trans. Amer. Math. Soc. 110 (1964), 493--518. 159802
1964
-
[25]
Shan, Dynamical localization for finitely differentiable quasi-periodic long-range operators, J
Y. Shan, Dynamical localization for finitely differentiable quasi-periodic long-range operators, J. Differential Equations 427 (2025), 803--826. 4862298
2025
-
[26]
Shamis and S
M. Shamis and S. Sodin, Upper bounds on quantum dynamics in arbitrary dimension, J. Funct. Anal. 285 (2023), no. 7, Paper No. 110034, 20. 4604835
2023
-
[27]
Yomdin, C^k -resolution of semialgebraic mappings
Y. Yomdin, C^k -resolution of semialgebraic mappings. A ddendum to: `` V olume growth and entropy'' , Israel J. Math. 57 (1987), no. 3, 301--317. 889980
1987
-
[28]
Young, Ballistic transport for limit-periodic S chr\"odinger operators in one dimension , J
G. Young, Ballistic transport for limit-periodic S chr\"odinger operators in one dimension , J. Spectr. Theory 13 (2023), no. 2, 451--489. 4651730
2023
-
[29]
Zhao, Ballistic transport in one-dimensional quasi-periodic continuous S chr\" o dinger equation , J
Z. Zhao, Ballistic transport in one-dimensional quasi-periodic continuous S chr\" o dinger equation , J. Differ. Equ. 262 (2017), no. 9, 4523--4566. 3608164
2017
-
[30]
Zhang and Z
Z. Zhang and Z. Zhao, Ballistic transport and absolute continuity of one-frequency S chr\" o dinger operators , Comm. Math. Phys. 351 (2017), no. 3, 877--921. 3623240
2017
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.