Interplay of localization and topology in disordered dimerized array of Rydberg atoms

Basic Information

• Paper ID: 2505.07720
• Title: Interplay of localization and topology in disordered dimerized array of Rydberg atoms
• Authors: Maksym Prodius, Adith Sai Aramthottil, Jakub Zakrzewski
• Affiliation: Jagiellonian University, Kraków, Poland
• Classification: cond-mat.dis-nn, cond-mat.quant-gas, quant-ph
• Publication Date: November 26, 2025
• Paper Link: https://arxiv.org/abs/2505.07720v2

Abstract

This paper investigates the effects of positional disorder and dimerization on the properties of excited states in Rydberg atom optical tweezers arrays. The studied system is a one-dimensional spin-1/2 Hamiltonian with long-range power-law decaying tunneling. Through numerical simulations, it is found that under parameter space constraints, the localized phase induced by disorder and dimerization mechanisms differs from standard many-body localization (MBL), instead arising from a collection of Hilbert space fragmentation (HSF) induced by small spin spacings. Despite partial spin-glass order, the system carries abundant symmetry-protected topological (SPT) states throughout the entire energy spectrum.

Research Background and Motivation

Research Questions

1. Core Problem: How do positional disorder and dimerization jointly influence the localization properties and topological characteristics of Rydberg atom arrays with long-range interactions?
2. Key Challenge: Understanding the coexistence relationship between non-standard many-body localization mechanisms and symmetry-protected topological order

Research Significance

1. Experimental Feasibility: Rydberg atom optical tweezers arrays provide a controllable experimental platform enabling precise manipulation of atomic positions and interactions
2. Theoretical Importance: Exploring novel non-ergodic mechanisms beyond the standard MBL paradigm
3. Topological Physics: Investigating the existence of SPT order in excited states, which is crucial for understanding topological protection mechanisms

Limitations of Existing Methods

1. Standard MBL Theory: The conventional MBL theory based on local integrals of motion (LIOMs) is inapplicable to systems with bond disorder and long-range interactions
2. Topological State Detection: Traditional entanglement spectrum degeneracy fails to correctly identify topological properties in the presence of cat states
3. Experimental Limitations: Most theoretical studies focus on ground states, with insufficient research on topological properties of excited states

Research Motivation

By combining real-space renormalization group (RSRG-X) methods with numerical exact diagonalization, this work systematically investigates the complete spectral properties of disordered dimerized chains, revealing the interplay between Hilbert space fragmentation, spin-glass order, and topological order.

Core Contributions

1. Revealing Non-Standard Localization Mechanisms: Demonstrating that the localized phase induced by positional disorder and dimerization is not standard MBL, but rather a collection of Hilbert space fragmentation caused by spin pairing
2. Discovering Spin-Glass Order: Observing finite Edwards-Anderson spin-glass order parameters in the pure XY model without ZZ interaction terms
3. Establishing RSRG-X Theoretical Framework: Proposing a real-space renormalization group scheme applicable to dimerized long-range interacting systems, successfully predicting eigenstate structure
4. Proposing Improved Topological Detection Method: Introducing a new disconnected entropy definition $S_2^D$, more effectively identifying topologically non-trivial states in excited states
5. Predicting Experimental Observables: Proposing experimental schemes to verify excited-state topological properties through time evolution of specific initial states, with oscillation periods scaling as $L^3$ and $L^6$ respectively

Detailed Methods

Task Definition

Studying the properties of one-dimensional long-range XY model under positional disorder and dimerization:

• Input: Disorder strength $W$ and dimerization parameter $\delta$
• Output: Localization properties, topological characteristics, and spin-glass order of eigenstates
• Constraints: Open boundary conditions, preserving $U(1) \times Z_2^T$ symmetry

Model Architecture

Hamiltonian

$H = J\sum_{i>j} \frac{1}{r_{ij}^3}(s_i^x s_j^x + s_i^y s_j^y)$

where spin positions are defined as: $r_i = \left(i + \delta(-1)^{i+1} - \frac{1}{2}\right)a + iD_{rb} + W_i$

• $\delta \in [-1, 1]$: dimerization parameter
• $W_i \in [-W, W]$: uniformly distributed positional disorder
• $D_{rb} = 0.2$: Rydberg blockade diameter
• $a = 1$: average lattice spacing

Parameter Space Constraints

Geometric constraint conditions: $W < a \cdot \min\left(\frac{1-\delta}{2}, \frac{\delta+1}{2}\right)$

forming a triangular parameter space as shown in Figure 1.

Technical Innovations

1. RSRG-X Scheme Adaptation

Improvements for dimerized systems:

• Strong Bond Identification: Prioritizing elimination of two-site operators corresponding to the strongest coupling $J_{ij} = J/r_{ij}^3$
• Ground State Selection Strategy:
  • For $|\pm\rangle_{ij} = \frac{1}{\sqrt{2}}(|\uparrow\downarrow\rangle \pm |\downarrow\uparrow\rangle)$ states: generating effective tunneling
  • For $|Z^\pm\rangle_{ij} = |\uparrow\uparrow\rangle/|\downarrow\downarrow\rangle$ states: generating chemical potential terms or pairing transition terms
• Path Dependence: Different elimination paths $\mathcal{P}$ correspond to different approximate eigenstates

2. Improved Disconnected Entropy Definition

Proposing $S_2^D$ definition (Figure 9b): $S_2^D = S_A^{ent} + S_B^{ent} - S_{A\cup B}^{ent} - S_{A\cap B}^{ent}$

where:

• $A = \{1, ..., L/2 + (L/2 \mod 2)\}$
• $B = \{2, ..., L/2 + (L/2 \mod 2), L\}$

Advantages:

• For $\delta < 0$: $S_2^D = 2$ indicates topologically non-trivial states, $S_2^D = 1$ indicates bulk domain wall states
• For $\delta > 0$: $S_2^D = 0$ indicates topologically trivial states
• Avoids false positives of traditional $S_1^D$ definition

3. Quantized State Fraction Statistics

Defining quantized state fraction: $P_Q(S^D = k) = \langle N_k/N \rangle$

where $N_k$ is the number of states with $S^D \in [k-\epsilon, k+\epsilon]$ ($\epsilon = 0.05$), and $N$ is the Hilbert space dimension.

Experimental Setup

Numerical Methods

1. Exact Diagonalization (ED): $L \leq 16$
2. POLFED Algorithm: $L > 16$, polynomial filtered exact diagonalization
3. Symmetry Sectors: Zero magnetization, even parity sector

System Parameters

• Chain Length: $L = 10, 12, 14, 16, 18, 20$
• Disorder Realizations: 100-5000 (depending on system size)
• Spectral Sampling: Up to 1000 eigenvalues near spectral center

Evaluation Metrics

1. Average Gap Ratio

$\langle r_n \rangle = \left\langle \frac{\min\{\delta_k, \delta_{k+1}\}}{\max\{\delta_k, \delta_{k+1}\}} \right\rangle$

• GOE value: $r_{GOE} \approx 0.53$ (ergodic phase)
• Poisson value: $r_{Poi} = 0.389$ (standard localization)

2. Half-Chain Entanglement Entropy

$S_{ent}(L/2) = -\text{Tr}(\rho \log_2 \rho)$

normalized to Page value $S_P$.

3. Edwards-Anderson Spin-Glass Order Parameter

$m_{EA} = \frac{1}{L^2}\sum_n \sum_{i \neq j} \langle n|s_i^z s_j^z|n\rangle^2$

Auxiliary quantity: $\chi_{EA} = L \cdot m_{EA}$

Comparative Analysis

• Horizontal Slices: Fixed $\delta = 0$, varying $W$
• Vertical Slices: Fixed $W = 0.1$, varying $\delta$
• Boundary Scanning: Along triangular parameter space boundary

Experimental Results

Main Results

1. Phase Diagram Structure (Figure 2)

In the $(\delta, W)$ parameter space:

• Ergodic Region: Central low-$W$, $\delta > 0$ region, $\langle r_n \rangle \approx 0.53$, $S_{ent}/S_P \approx 1$
• Localized Region: Large $W$ or $|\delta|$ region, characterized by:
  • Upper region (large $W$ small $|\delta|$): $\langle r_n \rangle \approx 0.4$ (slightly above Poisson)
  • Side regions (large $|\delta|$): $\langle r_n \rangle < 0.389$ (strong sub-Poisson)

2. Disorder-Driven Localization Transition (Figures 3a-b)

Fixed $\delta = 0$:

• Crossing Point: $W_c \approx 0.33$
• Ergodic Side ($W < W_c$):
  • $\langle r_n \rangle \to 0.53$ ($L \geq 14$)
  • $S_{ent}/S_P$ exhibits volume-law scaling
• Localized Side ($W > W_c$):
  • $\langle r_n \rangle \approx 0.4$ (not reaching Poisson value)
  • $S_{ent}/S_P$ exhibits sub-volume scaling

3. Dimerization-Driven Transition (Figures 3c-d)

Fixed $W = 0.1$:

• Two Crossing Points: $\delta \approx -0.35$ and $\delta \approx 0.4$
• Anomalous Feature: Entanglement entropy curves in localized phase do not monotonically order by system size (due to cutting position on strong/weak bonds)
• Sub-Poisson Behavior: At large $|\delta|$, $\langle r_n \rangle$ significantly below 0.389

4. Spectral Fragmentation (Figure 6)

For $W = 0.5, \delta = 0.0$, energy density over 1000 disorder realizations:

• Overall Distribution: Lorentzian-like rather than Gaussian
• Individual Realizations:
  • Some exhibit Gaussian (small displacements or collective translations)
  • Some display 2-3 "islands" (fragmentation due to spin pairing)
• Physical Mechanism: Different disorder realizations produce different Hilbert space fragmentation patterns

Ablation Studies

1. Entanglement Entropy and Gap Ratio Distributions (Figure 5)

Detailed distributions at three typical points:

Point A ($W=0.1, \delta=-0.8$, lower left):

• Gap Ratio: Strong level attraction, resonance peak at $r_n \approx 0.45$
• Entanglement Entropy: Peaks at $S_{ent} = 1, 2$ (integer values)
• Interpretation: Cat states and edge-bulk exchange states

Point B ($W=0.5, \delta=0.0$, upper):

• Gap Ratio: Dip at small gaps (residual level repulsion)
• Entanglement Entropy: Multi-peak structure, $S_{ent} \gtrsim 1, 2$
• Interpretation: Frozen spin domains and maximally correlated subsystems

Point C ($W=0.1, \delta=0.8$, lower right):

• Gap Ratio: Similar level attraction and resonance peaks as Point A
• Entanglement Entropy: Main peak at $S_{ent} = 1, 2$
• Interpretation: Trivial topological dimer exchange

2. Spin-Glass Order Analysis (Figure 7)

• Vertical Slice ($\delta=0$):
  • $m_{EA}$ increases with $W$
  • $\chi_{EA}$ crosses at $W \approx 0.37$
  • Extrapolation suggests $m_{EA}(L\to\infty) \neq 0$
• Horizontal Slice ($W=0.1$):
  • Two $\chi_{EA}$ crossing points correspond to dimerization transitions
  • Finite $m_{EA}$ in large $|\delta|$ region
• Scaling Behavior: $\chi_{EA} \propto L$ in localized region, $\chi_{EA} \sim 1/L$ in ergodic region

3. Topological State Fractions (Figures 11, 13)

$\delta = -0.8$ (non-trivial phase):

• $P_Q(S_2^D = 2)$: Decreases from ~15% at $L=10$ to ~10% at $L=18$
• $P_Q(S_2^D = 1)$: Decreases from ~20% to ~15%
• Extrapolation (Figure 12): Finite-size fitting suggests non-zero asymptotic value

$\delta = 0.8$ (trivial phase):

• $P_Q(S_2^D = 0)$: Dominant, approximately 70-80%
• $P_Q(S_2^D = 1,2)$: Negligible

Boundary Scanning (Figure 13):

• Left boundary (negative $\delta$): Non-trivial state fraction significant at strong dimerization
• Right boundary (positive $\delta$): Almost entirely trivial states

Case Studies

Time Dynamics Verification (Figures 14-15)

Case 1: Edge Oscillations (Figure 14)

• Initial State: $|\Psi\rangle = |\uparrow\rangle_1 |+\rangle_{2,3} |-\rangle_{4,5} \cdots |\downarrow\rangle_L$
• Physical Picture: Superposition of two approximate eigenstates $|\pm\rangle_{1,L}$
• Observable: Edge magnetization $s_1^z, s_L^z$
• Results:
  • Period $T = 2\pi/(E_1 - E_2)$
  • RSRG-X prediction consistent with exact evolution
  • Disorder average: $T \propto L^3$ (confirmed by fitting)

Case 2: Edge-Bulk Exchange (Figure 15)

• Initial State: $|\uparrow\rangle_1 \otimes |-\rangle_{2,3} \otimes \cdots \otimes |Z^-\rangle_{L/2,L/2+1} \otimes \cdots \otimes |\uparrow\rangle_L$
• Physical Picture: Exchange between edge $|Z^+\rangle_{1,L}$ and bulk $|Z^-\rangle_{L/2,L/2+1}$
• Results:
  • Oscillation period $T \propto L^6$ (logarithmic fit: $\ln T = 6.17\ln L - 0.35$)
  • Energy scale: $\Delta E \sim J_{1,L/2+1}J_{L/2+1,L}/J_{L/2,L/2+1}$

Experimental Findings

1. Non-Standard Localization: Gap ratio behavior in the localized phase (sub-Poisson or slightly super-Poisson) differs significantly from standard MBL
2. Hilbert Space Fragmentation: "Island" structure in the spectrum indicates the system is a collection of fragmented Hamiltonians
3. Topological-Glass Coexistence: Despite spin-glass order, substantial SPT states remain (~25-30% at $\delta=-0.8$)
4. Integer Entanglement Entropy: Eigenstate entanglement entropy in localized region concentrates at integer values, reflecting cat states and frozen domain structures
5. RSRG-X Effectiveness: Renormalization group predictions of state structure and dynamics highly consistent with exact calculations

Related Work

Many-Body Localization

• Standard MBL: First proposed by Fleishman & Anderson (1980), based on LIOM theoretical framework 12-15
• Experimental Verification: Observed in optical lattices by Schreiber et al. (2015) 16, realized in ion traps by Smith et al. (2016) 17
• Distinction in This Work: Bond disorder rather than positional disorder, long-range interactions, non-local integrals of motion

Hilbert Space Fragmentation

• Tilted Systems: Rakovszky et al. (2020) studied statistical localization 24, dipole moment conservation causing fragmentation 23, 26-30
• Experimental Observation: Observed in superconducting quantum processors by Guo et al. (2021) 32, in tilted Fermi-Hubbard chains by Kohlert et al. (2023) 33
• Contribution in This Work: Disorder-induced fragmentation with different patterns for each realization

Topological Order and Localization

• Theoretical Proposal: Huse et al. (2013) proposed MBL can protect quantum order 35
• SPT in MBL: Chandran et al. (2014) studied symmetry-protected topological order 36, Vasseur et al. (2016) analyzed particle-hole symmetry 38
• Experimental Implementation: SPT phase observed in Rydberg atoms by de Léséleuc et al. (2019) 41
• Innovation in This Work: Excited-state SPT detection method, glass order coexistence analysis

Real-Space Renormalization Group

• SDRG: Strong disorder RG proposed by Dasgupta & Ma (1980) 63, applied to random antiferromagnetic chains by Fisher (1994) 64
• RSRG-X: Extended to excited states by Pekker et al. (2014) 50
• Bond Disorder Application: Applied to bond-disordered spin chains by Aramthottil et al. (2024) 48, extended to dimerized systems in this work

Disconnected Entropy

• Theory: Proposed by Zeng & Zhou (2016) for SPT detection 69
• Application: Used for one-dimensional topological superconductors by Fromholz et al. (2020) 71
• Improvement in This Work: New partition definition $S_2^D$, avoiding false positives

Conclusions and Discussion

Main Conclusions

1. Localization Mechanism: The localization induced by positional disorder and dimerization in Rydberg atom arrays is not standard MBL, but rather a manifestation of Hilbert space fragmentation collections
2. Phase Diagram Structure: In the $(\delta, W)$ parameter space, a central ergodic region is surrounded by localized regions, forming a triangular phase diagram
3. Spin-Glass Order: The pure XY model (without ZZ terms) exhibits finite Edwards-Anderson order parameters in the localized region
4. Topological State Existence: Despite partial glass-like spectrum, approximately 25-30% of excited states in the $\delta < 0$ region maintain topologically non-trivial characteristics
5. RSRG-X Applicability: Real-space renormalization group method successfully predicts eigenstate structure and time dynamics, demonstrating its effectiveness in long-range interacting systems

Limitations

1. System Size Constraints

• Maximum Chain Length: $L = 20$ (ED), $L = 18$ (POLFED full spectrum)
• Extrapolation Uncertainty: Thermodynamic limit behavior of topological state fractions requires larger system verification
• Finite-Size Effects: Non-monotonic entanglement entropy curves in Figure 3c reflect cutting position dependence

2. Parameter Space Constraints

• Geometric Constraint: Equation (3) limits explorable $W$-$\delta$ region
• Fixed Parameters: $a=1$, $D_{rb}=0.2$, different parameters may alter phase diagram
• One-Dimensional Limitation: Higher-dimensional systems may exhibit different physics

3. Numerical Method Limitations

• POLFED Accuracy: Only partial spectrum accessible for large systems
• Disorder Averaging: Even with 5000 realizations, statistical errors remain
• Symmetry Sectors: Only zero magnetization even parity sector studied

4. Theoretical Understanding

• Gap Ratio Resonance Peak: Peak at $r_n \approx 0.45$ lacks theoretical explanation
• Glass-Topology Interface: Microscopic mechanism of coexistence of two orders in same spectrum unclear
• Non-Local Integrals of Motion: Explicit forms not clearly constructed

Future Directions

1. Experimental Verification

• Rydberg Platform: Implementing controllable positional disorder using optical tweezers arrays
• Dynamics Observation: Measuring predicted edge oscillations ($T \propto L^3, L^6$)
• Topological Detection: Measuring disconnected entropy through many-body interference

2. Theoretical Extensions

• Two-Dimensional Systems: Studying topological phase transitions in planar arrays
• Other Interactions: Considering van der Waals terms (Appendix B shows $\Delta \sim 0.1$ as critical point)
• Finite Temperature: Investigating thermal excitation effects on topological order

3. Numerical Improvements

• Larger Systems: Developing tensor network methods for $L > 20$
• Full Spectrum Analysis: Improving algorithms for complete spectral access
• Other Symmetry Sectors: Verifying four-fold degeneracy

4. Physical Mechanisms

• Integral of Motion Construction: Explicitly constructing non-local integrals of motion
• Glass State Microstructure: Understanding spin-glass order origins
• Resonance Peak Explanation: Developing theory explaining gap ratio resonance features

In-Depth Evaluation

Strengths

1. Methodological Innovation

• RSRG-X Adaptation: First systematic application of real-space renormalization group to dimerized long-range systems, with ground state selection strategy ($|\pm\rangle$ vs $|Z^\pm\rangle$) reflecting deep physical insight
• Improved Topological Detection: $S_2^D$ definition cleverly resolves cat state interference, partition design highly targeted
• Multi-Method Integration: ED, POLFED, RSRG-X, and statistical analysis form complete toolchain

2. Experimental Sufficiency

• Full Parameter Space Coverage: Triangular sweep in Figure 2, orthogonal slices in Figure 3, boundary analysis in Figures 4/13 provide comprehensive picture
• Multi-Scale Verification: System size scaling $L = 10-20$ rigorously analyzed
• Statistical Reliability: 100-5000 disorder realizations ensure statistical significance
• Multi-Angle Characterization: Gap ratio, entanglement entropy, spin-glass order parameter, disconnected entropy, time dynamics mutually corroborate

3. Result Convincingness

• Quantitative Predictions: Precise verification of time dynamics $T \propto L^3, L^6$ (Figures 14-15)
• Theoretical Self-Consistency: RSRG-X predictions highly consistent with numerical results
• Novel Physical Discoveries:
  • Sub-Poisson gap ratios (Figure 3d, 5a,c)
  • Integer entanglement entropy distributions (Figure 5d-f)
  • Spectral fragmentation (Figure 6)
  • Spin-glass order without ZZ terms (Figure 7)

4. Writing Clarity

• Logical Structure: Phase diagram → statistical properties → glass order → RG theory → topological analysis → dynamics verification, progressive development
• Figure Quality: 22 main figures + appendix figures, high information density and readability
• Physical Imagery: RSRG-X scheme illustration (Figure 8), partition definition (Figure 9) clear and intuitive
• Literature Review: Comprehensive introduction covering MBL, fragmentation, SPT fields

Weaknesses

1. Method Limitations

• RSRG-X Approximation: Path dependence insufficiently discussed, error estimates for different paths missing
• Extrapolation Uncertainty: Linear fitting in Figure 12 over $L=10-18$ range, extrapolation to $L\to\infty$ may have systematic bias
• Symmetry Sectors: Only zero magnetization even parity analyzed, four-fold degeneracy verification costly (mentioned in Section 6 but not systematically executed)

2. Experimental Setup Defects

• Parameter Selection: Physical motivation for $D_{rb}=0.2$ insufficient, sensitivity analysis missing
• Spectral Sampling: "Up to 1000 eigenvalues near spectral center" may miss edge state information
• Disorder Distribution: Only uniform distribution $W_i \in [-W, W]$ considered, other distributions (e.g., Gaussian) unexplored

3. Analysis Insufficiency

• Resonance Peak: Gap ratio resonance peak at $r_n \approx 0.45$ in Figures 5a,c "currently unexplained" (end of Section 3), undermining theoretical completeness
• Glass-Topology Boundary: Which states are glass states, which are SPT states? How to distinguish? (Section 6 discussion vague)
• ZZ Term Impact: Appendix B shows phase transition near $\Delta=0.1$, but main text lacks deep physical mechanism discussion
• Finite-Size Trend: Figure 11c inset shows $P_Q(S_2^D=2)$ decreasing with $L$, but extrapolation conclusion not explicit

4. Expression Issues

• "Extensive fraction": Abstract and conclusions claim "extensive fraction of SPT states," but Figure 12 extrapolation incompletely supports this (may trend to zero as $L\to\infty$)
• "Does not resemble standard MBL": Quantitative comparison with standard MBL insufficient, only qualitative description
• Terminology Mixing: "Localized phase" sometimes means MBL, sometimes fragmentation, needs clearer distinction

Impact

1. Field Contributions

• Theory: Establishing RSRG-X framework for bond disorder + long-range interaction systems, generalizable to other platforms (polar molecules, ion traps)
• Experiment: Providing clear predictions for Rydberg atom experiments (dynamics oscillation periods), directly verifiable
• Concepts: Challenging traditional "MBL requires positional disorder" notion, expanding understanding of non-ergodic mechanisms

2. Practical Value

• Quantum Simulation: Guiding Rydberg platform design for topological quantum states
• Topological Detection: $S_2^D$ method applicable to other open-chain SPT systems
• Error Diagnosis: Understanding positional error impact on quantum computing

3. Reproducibility

• Code: Not provided (common issue), but methods described in detail
• Parameters: All hyperparameters explicit ($a=1$, $D_{rb}=0.2$, $\epsilon=0.05$, etc.)
• Data: Figure data points clear, error bars annotated (e.g., Figure 7 inset)

4. Limitations

• Computational Cost: POLFED algorithm high barrier for general researchers
• Experimental Difficulty: $L=18$ Rydberg array requires advanced techniques
• Theory Gap: Lacking analytical results, relying on numerics

Applicable Scenarios

1. Direct Application

• Rydberg Atom Arrays: Paper's primary platform, $1/r^3$ dipole interaction naturally realized
• Polar Molecules: Similar long-range interactions, tunable positions and dimerization
• Ion Traps: Programmable long-range Ising/XY models

2. Requiring Modification

• Short-Range Models: RSRG-X scheme needs adjustment (original Fisher method 64 more suitable)
• Higher Dimensions: Fragmentation patterns more complex, topological invariants need redefinition
• Finite Temperature: Current zero-temperature results, thermal excitation effects require separate study

3. Inapplicable

• Standard MBL Systems: Method targets bond disorder, positional disorder needs different treatment
• Gapless Bulk Systems: RSRG-X requires wide gap distribution, metallic phases unsuitable
• Strong Interaction Limit: $U \gg J$ requires superexchange perturbation theory

Potential Extensions

1. Machine Learning: Using neural networks to learn disorder realization → fragmentation pattern mapping
2. Quantum Control: Exploiting topological state robustness for quantum storage design
3. Many-Body Scars: Investigating correlation between fragmentation and quantum scars (references 18-22)
4. Open Systems: Adding dissipation, studying topological order stability

Selected Key References

1. MBL Basics: Nandkishore & Huse, Annu. Rev. Condens. Matter Phys. 6, 15 (2015) - Review
2. Hilbert Space Fragmentation: Rakovszky et al., Phys. Rev. B 101, 125126 (2020) - Statistical localization
3. SPT and MBL: Vasseur et al., Phys. Rev. B 93, 134207 (2016) - Particle-hole symmetry
4. Rydberg Experiments: de Léséleuc et al., Science 365, 775 (2019) - SPT phase observation
5. RSRG-X: Pekker et al., Phys. Rev. X 4, 011052 (2014) - Original method
6. Bond Disorder: Aramthottil et al., Phys. Rev. Lett. 133, 196302 (2024) - Prior work from same group
7. Disconnected Entropy: Zeng & Zhou, Europhys. Lett. 113, 56001 (2016) - Topological invariant
8. Spin Glass: Parisi, Phys. Rev. Lett. 50, 1946 (1983) - Edwards-Anderson parameter

Overall Assessment: This is a high-quality theoretical physics paper excelling in methodological innovation (RSRG-X adaptation, improved disconnected entropy), numerical rigor (multi-scale multi-metric analysis), and physical insight (fragmentation mechanism, glass-topology coexistence). Main contributions reveal non-standard localization mechanisms in Rydberg atom arrays and propose experimentally verifiable predictions. Shortcomings include finite-size extrapolation uncertainty and lack of theoretical explanation for certain phenomena (gap ratio resonance peaks). The paper significantly impacts both condensed matter theory (non-ergodicity, topological matter) and cold atom experimental fields, providing new paradigms for understanding long-range disordered systems. Recommended future work focuses on larger system simulations, experimental verification, and closing theoretical gaps.