Semi-fractality and localization on a chiral Cayley tree
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.
Kai: We're looking at the paper: "Semi-fractality and localization on a chiral Cayley tree".
Mira: Here's the gist of the paper. Semi-fractality and localization on a chiral Cayley tree studies how quantum particle hopping on an infinite Cayley tree with nearest-neighbor hopping amplitudes drawn from a distribution singular as t−a near weak links and no on-site disorder develops broad power-law tails in the local density of states, implying an unusual form of wave-function statistics called semi-fractality The extended and non-ergodic phases in singleparticle
1–3: and many-body localization
4–8: have generated renewed interest in the condensed matter community
8–14: .
Kai: The old paradigm stating that the only non-ergodic phase is the one in which eigenfunctions are exponentially localized in the real space (for a singleparticle localization) or in the Hilbert space (for manybody localization) has changed in light of progress made recently In fact, a handful of works demonstrated that the non-ergodic wave functions may be extended, exhibiting mono- or multi-fractal properties
9, 11, 12, 15–17: .
Mira: The corresponding measure that quantifies such properties is the set of fractal dimensions Dq = dSq/d ln N, where N is the Hilbert space dimension and Sq is the Rényi (q ≠ 1) or Shannon-von Neumann (q = 1) entropy associated with the wave function coefficients
18, 19: .
Kai: Defining Semi-Fractality The missing case in the classification of extended and non-ergodic phases is what they name semifractal phase, in which D1 = 1 but Dq β ≥ 1.
Title and authors: Kai: Let's talk about who did this work and what they are calling it. The paper is titled "Semi-fractality and localization on a chiral Cayley tree." It involves Carlo Vanoni, Vladimir Kravtsov, and Boris Altshuler.
Mira: They’re looking at how quantum particles move on an infinite Cayley tree, focusing specifically on the way the hopping strengths behave near weak links.
Lev: From an engineering standpoint, I want to know if this setup is even feasible to simulate accurately. Is this something we can realistically build a model for in a simulator?
Kai: It’s about modeling systems where you have disorder that's singular, meaning the distribution of those hopping amplitudes doesn't behave nicely at small values—specifically, near zero.
Mira: The authors use population dynamics to solve the cavity equations for the propagator on this graph and show that this process results in broad power-law tails in the local density of states.
Lev: Power-law tails are usually a sign of criticality or scale invariance, but here it’s linked directly to the geometry of the tree and that singular hopping distribution.
Kai: The authors argue that these broad power-law tails imply wave function statistics they call semi-fractality, which is a hybrid state where the eigenstates span a huge part of the system.
Mira: But crucially, while they span an extensive fraction of the system, their higher moments are behaving in a way characteristic of a multifractal state.
The paper's summary: Kai: So to summarize what they found in "Semi-fractality and localization on a chiral Cayley tree," it’s about identifying this specific hybrid phase called semi-fractality.
Mira: They define this phase by the fractal dimensions Dq. They find that in this case, D1 is equal to one, which is what you'd expect for an ergodic system where everything spreads out, but the higher moments Dq for q greater than one are greater than or equal to one.
Lev: So they’re describing a state that’s spatially extensive but statistically complex in terms of its scaling properties. That means the localization isn't simple exponential decay anymore.
Kai: Right, and what makes this phase unique is that the fractal dimension Dq must be non-analytic, meaning it has a jump in its derivative when you move from one value of q to another, specifically at a point called beta greater than one.
Mira: The paper attributes this behavior to the chiral symmetry inherent in the infinite Cayley tree structure itself, which strongly influences how these wave functions behave at the center of the spectrum.
Lev: So it’s not just about the disorder distribution; it’s about how that disorder interacts with a specific geometric symmetry—the chirality—to create this new statistical regime.
The paper's improvements: Kai: The authors suggest a few ways to look at this problem and build on their findings, especially regarding how we can predict these phases better.
Mira: They point out that they’ve distinguished between two main regimes based on the parameter beta. If beta is greater than one, you get a semi-fractal phase where Dq equals one for q equal to one, but Dq beta shows multifractal behavior for higher q values.
Lev: And if we look at the other side, where beta is less than one, they describe a localized state where the fractal dimension D1 is zero. That’s a clear distinction between the two regimes they are mapping out.
Kai: They also connect this to how the eigenstate weights behave microscopically. They show that there’s a power-law hierarchy in these weights, specifically something like w of r proportional to N minus one minus mu times r raised to the power of mu, where mu is between zero and one.
Mira: This hierarchy of weights leads directly to a piecewise-linear spectrum called tau q, which is the mathematical signature of semi-fractality they found. It’s a direct link between the microscopic weight structure and the macroscopic fractal dimension scaling.
Conclusion: Kai: So wrapping up on "Semi-fractality and localization on a chiral Cayley tree," the main point is that we have identified two new classes of eigenfunction statistics: semi-fractal and semi-localized.
Mira: The key implication is that these phases are directly tied to the chiral symmetry of the underlying Hamiltonian, suggesting that an exact or approximate chiral structure exists in these systems.
Lev: From a hardware perspective, this means when we try to run simulations on real machines, we need to account for this specific statistical behavior of the wave functions if we want accurate results.
Kai: It suggests that even in what looks like a standard localization problem, there are deeper structural properties dictated by the geometry and symmetry of the underlying lattice.
Mira: This work opens up a new way to categorize non-ergodic phases beyond just exponential decay, giving us this semi-fractal regime where the state is both extensive and statistically complex.
Lev: It’s interesting because they mention that this connection to chirality is supported by comparing results from exact diagonalization and population dynamics, noting a discrepancy in the value of beta between those two methods.
Kai: So, despite the complexity, it seems like chiral symmetry might be a fundamental driver for these unusual statistical behaviors in quantum systems.
Department of Physics, Princeton University · The Abdus Salam International Centre for Theoretical Physics Institute of Trieste Department of Physics, Columbia University
cond-mat.dis-nn, cond-mat.stat-mech, quant-ph
Submitted: 2026-07-20
Updated: 2026-11-09
Comments: 13 + 4 pages, 13 + 4 figures. Added analytical theory section. Comments are welcome!
Code: https://github.com/CarloVanoni/Chiral_Cayley_Tree
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 90/100
The gist: Semi-fractality and localization on a chiral Cayley tree studies how quantum particle hopping on an infinite Cayley tree with nearest-neighbor hopping amplitudes drawn from a distribution singular as…
Key concepts
- Semi-fractality
- A classification of non-ergodic phases where wave functions have a specific fractal dimension profile. In this phase, the dimension Dq equals 1 for q=1 (like an ergodic system), but other dimensions Dq for q greater than 1 are greater than or equal to 1, indicating a hybrid behavior.
- Chiral Symmetry
- A property of the Hamiltonian that is either explicit or hidden, which is crucial in this study. The paper suggests that this symmetry is responsible for the observed semi-fractal phase and the specific power-law distributions found in the local density of states.
- Local Density of States (LDoS)
- This measures how many quantum energy levels are available at a specific spatial location on the tree. The paper analyzes its distribution, P(ρ), which shows power-law tails. These tails are key indicators that connect the hopping disorder to the system's fractal dimensions.
- Power-law Hierarchy of Eigenstate Weights
- This refers to a pattern in how the probabilities of different quantum states are ranked. The paper suggests this hierarchy, which is conjectured to stem from chirality, creates a piecewise-linear spectrum characteristic of the semi-fractal state.
Terminology
Summary
Semi-fractality and localization on a chiral Cayley tree studies how quantum particle hopping on an infinite Cayley tree with nearest-neighbor hopping amplitudes drawn from a distribution singular as t−a near weak links and no on-site disorder develops broad power-law tails in the local density of states, implying an unusual form of wave-function statistics called semi-fractality
Introduction to Non-Ergodic Phases
The extended and non-ergodic phases in singleparticle [1–3] and many-body localization [4–8] have generated renewed interest in the condensed matter community [8–14]. The old paradigm stating that the only non-ergodic phase is the one in which eigenfunctions are exponentially localized in the real space (for a singleparticle localization) or in the Hilbert space (for manybody localization) has changed in light of progress made recently In fact, a handful of works demonstrated that the non-ergodic wave functions may be extended, exhibiting mono- or multi-fractal properties [9, 11, 12, 15–17]. p. 0 The corresponding measure that quantifies such properties is the set of fractal dimensions Dq = dSq/d ln N, where N is the Hilbert space dimension and Sq is the Rényi (q ≠ 1) or Shannon-von Neumann (q = 1) entropy associated with the wave function coefficients [18, 19]. p. 0
Defining Semi-Fractality
The missing case in the classification of extended and non-ergodic phases is what they name semifractal phase, in which D1 = 1 but Dq β ≥ 1. The support set of wave functions in such a phase is proportional to N, but other important measures, e. p. 0g., the return probability, R(t), which is determined by the fractal dimension D2, may show a typical multifractal behavior Rt ∝ t−D2 [3, 24, 25]. p. 0 According to its definition, Dq in the semi-fractal phase must be non-analytic, generally with a jump of the derivative dDq/dq at q = β > 1. p. 0
Chiral Symmetry and Power Laws
The paper presents evidence that the chiral symmetry (explicit or hidden, exact or approximate) is responsible for the semi-fractal phase observed in Refs. [26, 27]. The main finding is that in both systems (infinite Cayley tree and finite RRG) the distribution P(ρ) of the local density of states (LDoS) ρ(E; r) = Pnψn(r)2 δ(E−En) behaves as a power-law both at small and large ρ, with different exponents. p. 1 The presence of the power-law regions in P(ρ) leads to the linear segment in f(α) and thus to a non-analytic behavior of fractal dimensions Dq and to the semi-fractal phase in a certain region of parameters of the link disorder distribution. p. 1
Regimes of Semi-Fractality
The analysis distinguishes between two regimes based on the parameter β:
-
Case β > 1: This leads to a semi-fractal phase where Dq = 1 for q 1, as in an ergodic system, whereas Dq β, as in a multifractal system. p. 9 At the limiting value β = 1, the semi-localized state has Dq = 1 for q 1, thus combining ergodic and localized behavior. p. 9 p. 9
-
Case β < 1: This corresponds to a localized state where Dq = (N−q)/(N(1−β)) for q β, since D1 = 0. p. 4
Microscopic Interpretation of Eigenstate Weights
The paper shows that the semi-fractality is connected to the emergence of a power-law hierarchy in the eigenstate weights, which is conjectured to originate from chirality. The rank-ordered eigenstate probabilities exhibit a broad power-law regime, ⟨w(r)⟩ ∼ N−(1−µ)r µ, 0 < µ < 1. p. 8 This power-law hierarchy produces a piecewise-linear spectrum τ (q), characteristic of semi-fractality. p. 9
Experimental and Theoretical Connections
The chiral symmetry is suggested by the behavior of the LDoS distribution P(ρ) which consists of two power-law segments with exponents constrained by the Mirlin-Fyodorov symmetry. The connection to chirality is further supported by comparing results from exact diagonalization and population dynamics, where the discrepancy in the value of β between these methods suggests that the increased value of β should be used in Eq. p. 7 (24) to find Dq and f(α) for β < 1. p. 0
Conclusion on Symmetry
The main result is the identification of two new classes of eigenfunction statistics, termed semifractal and semi-localized, and their connection to the chiral symmetry of the Hamiltonian. The semi-fractal regime exhibits an unusual hybrid behavior: Dq = 1 for q 1, as in an ergodic system, whereas Dq β, as in a multifractal system. p. 9 At the limiting value β = 1, the semi-localized state has Dq = 1 for q 1, thus combining ergodic and localized behavior. The authors argue that an exact or approximate chiral structure is present in each of these examples. p. 9
Acknowledgements and Data
VEK is grateful to M. V. Feigel’man and I. M. Khaymovich for multiple illuminating discussions. The authors are grateful to A. p. 1 Lunkin for useful comments. Some of the numerical analysis and plotting code used in this work was developed with assistance from OpenAI Codex. p. 7 The code is made available in the GitHub repository in Ref. p. 0 [47].
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI can predict whether a quantum system, modeled on an infinite Cayley tree with nearest-neighbor hopping amplitudes drawn from a specific power-law distribution, will exhibit a semi-fractal or localized phase based on the exponent controlling that distribution. This is achieved by analyzing how the parameter a in the hopping distribution affects the fractal dimensions Dq and determines whether Dq jumps from 1 to 0 at q = 1.
-
Improved AI can characterize eigenfunction statistics using rank-ordered eigenstate probabilities to infer spectral properties of fractal dimensions. The system can use this information to distinguish between semi-fractal and semi-localized states by checking if the hierarchy of weights reproduces the piecewise-linear participation spectrum derived in Sec III.
-
Improved AI can analyze the distribution of local density of states (LDoS) and relate its power-law tails to multifractal spectra. The system can use this relationship, governed by Eq 14, to determine the exponent beta and gamma, which are functions of the disorder parameter a, thus providing a continuous mapping between hopping distribution parameters and eigenfunction statistics.
-
Improved AI can perform quantitative comparison between population dynamics simulations and exact diagonalization results for LDoS distributions on finite random regular graphs. The system can assess the accuracy of its theoretical predictions by comparing the power-law segments found in population dynamics against those obtained from exact diagonalization, noting that
the power laws in the latter are much more accurate.
Sources
- Emergent quantum chaos from correlations on a random graph
- Difference between level statistics, ergodicity and localization transitions on the Bethe lattice
- Wegner model in high dimension: U(1) symmetry breaking and a non-standard phase of disordered electronic matter, I. One-replica theory
- Support set of random wave-functions on the Bethe lattice
- Multifractal and Glassy Signatures of Non-Ergodic 2D Quantum Dynamics
- Novel Symmetry Classes in Mesoscopic Normal-Superconducting Hybrid Structures
Related papers
- Topological metal-insulator transitions in one-dimensional non-Hermitian quasicrystals: beyond PT-symmetry
- Microscopic and collective signatures of feature learning in neural networks
- Synchronous and asynchronous Monte Carlo dynamics of mean-field spin glasses with rotationally invariant couplings: what they share and where they differ
- Reentrant Localization Length in a Quasiperiodic Thue-Morse Chain
- Beyond Ground States: Learning Many-Body Hamiltonians for p-Bit Invertible Arithmetic
- Linear equivalence of nonlinear recurrent neural networks