arxiv: 2604.18989 · v1 · submitted 2026-04-21 · 🧮 math.AP · math-ph· math.MP· math.PR

Recognition: unknown

Anderson Localization for the hierarchical Anderson-Bernoulli model on mathbb{Z}^d

Shihe Liu , Yunfeng Shi , Zhifei Zhang

Authors on Pith no claims yet

Pith reviewed 2026-05-10 02:39 UTC · model grok-4.3

classification 🧮 math.AP math-phmath.MPmath.PR

keywords Anderson localizationhierarchical modelBernoulli potentialSchrödinger operatorunique continuationrandom latticedisordered media

0 comments

The pith

The hierarchical Anderson-Bernoulli model exhibits Anderson localization on lattices of arbitrary dimension.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper proves Anderson localization for a random Schrödinger operator on the integer lattice in any dimension, where the potential is built from a deterministic geometric hierarchical pattern overlaid with independent Bernoulli random variables. The spectrum is pure point almost surely and eigenfunctions decay exponentially. This matters because it removes the usual dimensional restrictions that limit many localization proofs, showing that a specific scale-by-scale structure in the disorder suffices to prevent wave propagation. The same method yields a probabilistic unique continuation result that bounds how far a solution can be controlled from partial data on the lattice.

Core claim

For the hierarchical Anderson-Bernoulli model on Z^d with arbitrary d, the random Schrödinger operator has almost surely pure point spectrum consisting of exponentially localized eigenfunctions.

What carries the argument

The hierarchical Anderson-Bernoulli potential, whose geometric self-similar structure combined with i.i.d. Bernoulli fluctuations permits inductive control across scales.

If this is right

Localization holds for arbitrary lattice dimension.
Eigenfunctions decay exponentially at a rate independent of dimension.
The spectrum is pure point almost surely.
A probabilistic unique continuation principle holds on the lattice.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same hierarchical construction may serve as a test case for localization proofs in other disordered systems where standard methods break down at high dimension.
Numerical diagonalization on large finite lattices with this potential could check the predicted localization length.
The unique continuation statement suggests new ways to reconstruct solutions from boundary data in random media.

Load-bearing premise

The potential must follow a precise geometric hierarchical pattern that interacts with the Bernoulli randomness to permit scale-by-scale analysis.

What would settle it

An explicit high-dimensional sample of the potential where some eigenfunction fails to decay exponentially or where the spectrum contains a continuous component would disprove the claim.

Figures

Figures reproduced from arXiv: 2604.18989 by Shihe Liu, Yunfeng Shi, Zhifei Zhang.

**Figure 1.** Figure 1: Hierarchical structure with well density N ≤ 5. Consider the deterministic Hamiltonian H0 = 2d − ∆ + Vhi. Denote by Spec(·) the spectrum of an operator. According to [JLMS85, Proposition 3.1], the spectrum of H0 satisfies Spec(H0) ∩ [h, ∞) = [h, h + 4d], Spec(H0) ∩ [0, h) ⊂ [0, 4d]. It was established in [JLMS85, Theorem 4.2] that, at the bottom of the spectrum (i.e., Spec(H0)∩ [0, h)), H0 (with a symmetri… view at source ↗

**Figure 2.** Figure 2: Visualization of the proof of Lemma 2.2. Regarding (2.7), if nj ∈ ∂ −B2, then the Dirichlet boundary condition ensures that nj+1 remains inside B2. Moreover, (2.7) implies 1 ≤ ⟨nj+1 − nj , ed⟩ ≤ 2, (2.9) which gives s ≤ ⟨ns − n0, ed⟩ ≤ L + ℓ. Consequently, |u(ns)| ≥ γ L+ℓ |u(n0)|. The inequality (2.9) together with the Dirichlet boundary condition also guarantees that the terminal point ns lies on ∂ −B2 in… view at source ↗

**Figure 3.** Figure 3: The event Ej1,j2,ℓ. From now on, we take condition probability on Vr,F . This leads to a cylindrical decomposition of the probability space: {0, 1} Λk = [ v∈{0,1}F {0, 1} S × {Vr,F = v} := [ v∈{0,1}F Cv. In this representation, both Euc and Ej1,j2,ℓ can be viewed as subsets of Cv. Recall the definition of Nuc given by (3.9). For i = 0, 1, let Ej1,j2,ℓ,i denote the event that Ej1,j2,ℓ and # n|ψj1 | ≥ exp{−… view at source ↗

**Figure 4.** Figure 4: Smallness generates from random walks. Now, noticing that the annulus Λk \Λ ′ k is non-resonant and Bj−1 \Bj , Bj \Bj are both subset of Λk \ Λ ′ k (therefore are both non-resonant), we can apply Lemma 3.1 to the expansion (4.13). Recall that 2d h − 2d − β − ι < 1 [PITH_FULL_IMAGE:figures/full_fig_p029_4.png] view at source ↗

**Figure 5.** Figure 5: The construction of the chain of sites in the martingale. for some constant q > 0. Clearly, the event XP > 1 10P represents that the strict monotonicity in (4.9) happens more than 1 10P times, and therefore (Lemma 4.3 tells us that deterministically nbj (Bj ) ≤ nbj−1(Bj−1)) nbP (BP ) ≤ max nb0(B0) − 1 10 P, 0 . (4.98) Now, recall that the nb0(B0) is the number of eigenvalues (counting multiplicities) o… view at source ↗

**Figure 6.** Figure 6: The iteration of the solution till the initial data. This fact also has an analogue for the tilted line T ′ k := {(n1, n2) ∈ Z 2 : n1 + n2 = k}. Based on the above discussions, we can construct the martingale in IL as follows. Choose the site X0 = 0 ∈ Z 2 and take the σ-algebra with B0 = {Ω, ∅} ⊂ F0 = σ({0}). Now, we already know the initial data u ≡ u0 on P0 ∪ P1. (Step 1) [PITH_FULL_IMAGE:figures/full_f… view at source ↗

**Figure 7.** Figure 7: Construction of L1. With this notation, we denote the region under L1 by R1, i.e., R1 =    T≤(X1)2−(X1)1 , if L1 = T(X1)2−(X1)1 , T ′ ≤(X1)1+(X1)2 , if L1 = T ′ (X1)1+(X1)2 . We construct the next filtration element by F1 = σ(F1), F1 := (R1 ∩ IL) ∪ B1. (A.5) Our construction ensures that B1 ⊂ F1, X1 ∈/ B1, X1 ∈ F1, X2 ∈/ F1. (A.6) Therefore, B1 ⊂ F1. If we further condition on F1, clearly the value of u… view at source ↗

**Figure 8.** Figure 8: Construction of the martingale (|u(Xj )|, Lj , ej , Xj+1, Zj ). We set the length of the martingale to be T = 1 10L. As in Section 4.5, we define Yj = Xj − 1 2 j = Z1 + Z2 + · · · + Zj − 1 2 j, 1 ≤ j ≤ T. (A.13) Then (Yj )1≤j≤T becomes a submartingale. Applying Azuma’s inequality yields P XT ≤ 1 10 T = P YT ≤ −( 1 2 − 1 10 )T ≤ exp{−T} = exp{− 1 10 L}. (A.14) Clearly, both the probabilistic estimat… view at source ↗

**Figure 9.** Figure 9: Illustration of the proof of Claim A.1 for d = 3. Finally, assume X1 ∈ Q L 100d (0). Then we have |X1 − x ′ |1 ≤ |X1|1 + |x ′ |1 ≤ 2|X1|1 ≤ L 50 ≤ k 10 . (A.23) That means X1 is also near the center of X(y0). Moreover, using (A.21) and (A.22) ensures that if we shift y0 by a vector m ∈ Z d with m1 + · · · + md−1 = 0, y0 + m ∈ IL, then we still have y0 + m ∈ e ′ 1 and exactly X(y0 + m) = X(y0) + m. Since (A… view at source ↗

read the original abstract

In this paper, we prove Anderson localization for a hierarchical Anderson-Bernoulli model on lattice with arbitrary dimension, where the potential is characterized by a geometric hierarchical structure combined with fluctuations induced by independent and identically distributed (i.i.d.) Bernoulli random variables. Our method is also applicable to proving a probabilistic unique continuation result on $\mathbb{Z}^d$.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper claims a proof of Anderson localization for this hierarchical Bernoulli model on Z^d in any dimension, which would be a real step forward if the details check out.

read the letter

The main takeaway is that this work proves Anderson localization for a hierarchical Anderson-Bernoulli model on the integer lattice in arbitrary dimension. The potential combines a fixed geometric hierarchy with i.i.d. Bernoulli fluctuations, and the same method yields a probabilistic unique continuation statement on Z^d. That combination lets them reach all dimensions, which is the part that stands out. Most localization arguments hit volume or decay problems when d grows, so the hierarchy probably supplies a recursive structure that keeps the estimates under control while the Bernoulli randomness supplies the necessary disorder. The abstract states the result cleanly and ties it to an application, which is useful. The model itself is concrete and the randomness is standard, so the setup does not look artificial. The soft spot is that the abstract gives no proof outline, so the multiscale analysis, resolvent bounds, and induction over hierarchy levels cannot be inspected. If the hierarchy is used only to simplify the geometry and the Bernoulli part handles the probabilistic estimates without extra fitting, the argument should hold; nothing visible suggests a circular step or a dimension-dependent gap. The paper is written for people already working on random operators and Anderson localization. A reader who cares about high-dimensional cases or structured potentials will see a concrete way to remove the usual dimension barrier. It deserves a serious referee because the claim is specific, the model is well-defined, and the result would matter inside the subfield even if it is not the most general possible. I would send it to peer review so the technical steps can be checked directly.

Referee Report

0 major / 0 minor

Summary. The manuscript proves Anderson localization for the hierarchical Anderson-Bernoulli model on the lattice Z^d in arbitrary dimension d. The potential is defined via a deterministic geometric hierarchical structure augmented by i.i.d. Bernoulli random fluctuations. The same method is applied to establish a probabilistic unique continuation principle on Z^d.

Significance. If the central proof holds, the result is of interest to the spectral theory community because it delivers localization in every dimension for a structured random potential where standard Anderson models remain open in low dimensions. The hierarchical construction evidently permits a controlled multiscale or inductive argument that bypasses some of the usual obstacles. The additional probabilistic unique-continuation statement is a useful byproduct that may find independent applications.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their careful reading of the manuscript, accurate summary of the results, and positive recommendation to accept. The report correctly identifies the interest of establishing Anderson localization in all dimensions for this structured random potential and notes the utility of the probabilistic unique continuation principle.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper claims a direct mathematical proof of Anderson localization for a hierarchical Anderson-Bernoulli model on Z^d (arbitrary dimension), defined via geometric hierarchical potential plus i.i.d. Bernoulli fluctuations, with an additional application to probabilistic unique continuation. No load-bearing steps reduce by construction to fitted parameters, self-definitions, or self-citation chains; the derivation chain is presented as self-contained against external mathematical benchmarks and does not invoke uniqueness theorems or ansatzes from prior author work in a circular manner. This matches the default expectation for a proof paper with independent content.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only the abstract is available, so the ledger is necessarily incomplete; no explicit free parameters, invented entities, or non-standard axioms are stated.

axioms (1)

domain assumption i.i.d. Bernoulli random variables are well-defined on the probability space
Invoked in the definition of the potential fluctuations.

pith-pipeline@v0.9.0 · 5352 in / 1028 out tokens · 34681 ms · 2026-05-10T02:39:05.104552+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

36 extracted references · 2 canonical work pages

[1]

Aizenman and S

M. Aizenman and S. Molchanov. Localization at large disorder and at extreme energies: an elementary derivation. Comm. Math. Phys. , 157(2):245--278, 1993

1993
[2]

P. W. Anderson. Absence of diffusion in certain random lattices. Phys. Rev. , 109(5):1492--1505, 1958

1958
[3]

Aizenman and S

M. Aizenman and S. Warzel. Random operators , volume 168 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2015. Disorder effects on quantum spectra and dynamics

2015
[4]

Bourgain, M

J. Bourgain, M. Goldstein, and W. Schlag. Anderson localization for S chr\" o dinger operators on Z^2 with quasi-periodic potential. Acta Math. , 188(1):41--86, 2002

2002
[5]

Bourgain and C

J. Bourgain and C. E. Kenig. On localization in the continuous A nderson- B ernoulli model in higher dimension. Invent. Math. , 161(2):389--426, 2005

2005
[6]

Buhovsky, A

L. Buhovsky, A. Logunov, E. Malinnikova, and M. Sodin. A discrete harmonic function bounded on a large portion of Z^2 is constant. Duke Math. J. , 171(6):1349--1378, 2022

2022
[7]

Bourgain

J. Bourgain. Random lattice S chr\" o dinger operators with decaying potential: some higher dimensional phenomena. In Geometric aspects of functional analysis , volume 1807 of Lecture Notes in Math. , pages 70--98. Springer, Berlin, 2003

2003
[8]

Bourgain

J. Bourgain. On localization for lattice S chr\" o dinger operators involving B ernoulli variables. In Geometric aspects of functional analysis , volume 1850 of Lecture Notes in Math. , pages 77--99. Springer, Berlin, 2004

2004
[9]

Bourgain

J. Bourgain. Anderson localization for quasi-periodic lattice S chr\" o dinger operators on Z^d , d arbitrary. Geom. Funct. Anal. , 17(3):682--706, 2007

2007
[10]

Carmona, A

R. Carmona, A. Klein, and F. Martinelli. Anderson localization for B ernoulli and other singular potentials. Comm. Math. Phys. , 108(1):41--66, 1987

1987
[11]

H. Cao, Y. Shi, and Z. Zhang. Localization and regularity of the integrated density of states for S chr\" o dinger operators on Z^d with C^2 -cosine like quasi-periodic potential. Comm. Math. Phys. , 404(1):495--561, 2023

2023
[12]

Delyon, Y

F. Delyon, Y. L\' e vy, and B. Souillard. Anderson localization for multidimensional systems at large disorder or large energy. Comm. Math. Phys. , 100(4):463--470, 1985

1985
[13]

Ding and C

J. Ding and C. K. Smart. Localization near the edge for the A nderson B ernoulli model on the two dimensional lattice. Invent. Math. , 219(2):467--506, 2020

2020
[14]

Damanik, R

D. Damanik, R. Sims, and G. Stolz. Localization for one-dimensional, continuum, B ernoulli- A nderson models. Duke Math. J. , 114(1):59--100, 2002

2002
[15]

Fr\" o hlich, F

J. Fr\" o hlich, F. Martinelli, E. Scoppola, and T. Spencer. Constructive proof of localization in the A nderson tight binding model. Comm. Math. Phys. , 101(1):21--46, 1985

1985
[16]

Fr\" o hlich and T

J. Fr\" o hlich and T. Spencer. Absence of diffusion in the A nderson tight binding model for large disorder or low energy. Comm. Math. Phys. , 88(2):151--184, 1983

1983
[17]

o hlich, T. Spencer, and P. Wittwer. Localization for a class of one-dimensional quasi-periodic S chr\

J. Fr\" o hlich, T. Spencer, and P. Wittwer. Localization for a class of one-dimensional quasi-periodic S chr\" o dinger operators. Comm. Math. Phys. , 132(1):5--25, 1990

1990
[18]

Germinet and A

F. Germinet and A. Klein. A comprehensive proof of localization for continuous A nderson models with singular random potentials. J. Eur. Math. Soc. (JEMS) , 15(1):53--143, 2013

2013
[19]

J. Z. Imbrie and R. Mavi. Level spacing for non-monotone A nderson models. J. Stat. Phys. , 162(6):1451--1484, 2016

2016
[20]

J. Z. Imbrie. Localization and eigenvalue statistics for the lattice A nderson model with discrete disorder. Rev. Math. Phys. , 33(8):Paper No. 2150024, 50, 2021

2021
[21]

Jitomirskaya

S. Jitomirskaya. Ergodic S chr\" o dinger operators (on one foot). In Spectral theory and mathematical physics: a F estschrift in honor of B arry S imon's 60th birthday , volume 76 of Proc. Sympos. Pure Math. , pages 613--647. Amer. Math. Soc., Providence, RI, 2007

2007
[22]

Jona-Lasinio, F

G. Jona-Lasinio, F. Martinelli, and E. Scoppola. New approach to the semiclassical limit of quantum mechanics. I . M ultiple tunnelings in one dimension. Comm. Math. Phys. , 80(2):223--254, 1981

1981
[23]

Jona-Lasinio, F

G. Jona-Lasinio, F. Martinelli, and E. Scoppola. Quantum particle in a hierarchical potential with tunnelling over arbitrarily large scales. J. Phys. A , 17(12):L635--L638, 1984

1984
[24]

Jona-Lasinio, F

G. Jona-Lasinio, F. Martinelli, and E. Scoppola. Multiple tunnelings in d dimensions: a quantum particle in a hierarchical potential. Ann. Inst. H. Poincar\' e Phys. Th\' e or. , 42(1):73--108, 1985

1985
[25]

o dinger operators. In Random S chr\

W. Kirsch. An invitation to random S chr\" o dinger operators. In Random S chr\" o dinger operators , volume 25 of Panor. Synth\`eses , pages 1--119. Soc. Math. France, Paris, 2008. With an appendix by Fr\' e d\' e ric Klopp

2008
[26]

Kritchevski

E. Kritchevski. Hierarchical A nderson model. 2008

2008
[27]

L. Li. Anderson- B ernoulli localization at large disorder on the 2 D lattice. Comm. Math. Phys. , 393(1):151--214, 2022

2022
[28]

S. Liu, Y. Shi, and Z. Zhang. Extended states for the random S chr \"o dinger operator on Z ^d\ (d 5) with decaying B ernoulli potential. arXiv:2505.04077 , 2025

work page arXiv 2025
[29]

S. Liu, Y. Shi, and Z. Zhang. On localization for the alloy-type A nderson- B ernoulli model with long-range hopping. arXiv:2508.12714 , 2025

work page arXiv 2025
[30]

Li and L

L. Li and L. Zhang. Anderson- B ernoulli localization on the three-dimensional lattice and discrete unique continuation principle. Duke Math. J. , 171(2):327--415, 2022

2022
[31]

Martinelli and E

F. Martinelli and E. Scoppola. Remark on the absence of absolutely continuous spectrum for d -dimensional S chr\" o dinger operators with random potential for large disorder or low energy. Comm. Math. Phys. , 97(3):465--471, 1985

1985
[32]

Martinelli and E

F. Martinelli and E. Scoppola. Introduction to the mathematical theory of A nderson localization. Riv. Nuovo Cimento (3) , 10(10):1--90, 1987

1987
[33]

Shubin, R

C. Shubin, R. Vakilian, and T. Wolff. Some harmonic analysis questions suggested by A nderson- B ernoulli models. Geom. Funct. Anal. , 8(5):932--964, 1998

1998
[34]

Simon and T

B. Simon and T. Wolff. Singular continuous spectrum under rank one perturbations and localization for random H amiltonians. Comm. Pure Appl. Math. , 39(1):75--90, 1986

1986
[35]

F. Wegner. Bounds on the density of states in disordered systems. Z. Phys. B , 44(1-2):9--15, 1981

1981
[36]

Renormalization group and dynamical maps for the hierarchical tight-binding problem

D W \"u rtz, T Schneider, A Politi, and M Zannetti. Renormalization group and dynamical maps for the hierarchical tight-binding problem. Physical Review B , 39(11):7829, 1989

1989