A derivation of the late-time volume law for local operator entanglement

arXiv:2603.25387 · quant-ph, cond-mat.other · Submitted 2026-03-26 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: Today's paper: "A derivation of the late-time volume law for local operator entanglement".

Mira: Local Operator Entanglement (LOE) serves as an indicator of quantum chaos in many-body systems,

Kai: First, who's behind it and why it matters.

Title and authors: Kai: So we’re starting by looking at the title and who wrote this paper, "A derivation of the late-time volume law for local operator entanglement," and it seems to involve John Goold, Guilherme Il´ario Correr, and Marco Cattaneo.

Mira: The authors are clearly experts in different areas—Goold with his work likely touching on the foundational aspects of quantum dynamics or perhaps spectral theory, while Correr and Cattaneo contribute the specific analytical machinery for this entanglement derivation.

Lev: I've seen papers involving spectral deconvolution and Majorana zero modes recently; I’m curious if their approach to deriving this scaling has any parallels with how we handle complex in-gap states in topological superconductors.

Kai: That’s a fair comparison, Lev, because both involve trying to extract meaningful physics from complex spectral information where the standard methods struggle.

Mira: The paper suggests that understanding the late-time behavior of LOE is crucial because it bridges the gap between what we observe numerically in chaotic dynamics and having a complete analytical picture of how operator entanglement actually grows.

Lev: If this derivation holds up, it could simplify our theoretical efforts when designing experiments to probe quantum chaos, potentially helping us focus on the most sensitive observables.

The paper's summary: Kai: Now we move into a summary of what the paper actually claims about the research; they’re focusing on deriving an explicit formula for late-time LOE that follows a volume law in chaotic systems.

Mira: Essentially, they set out to find an analytical expression for this scaling, and their derivation hinges on three main assumptions: first, a higher-order non-resonance condition for the Hamiltonian's eigenenergies; second, applying the Eigenstate Thermalization Hypothesis or ETH to the matrix elements of local operators; and third, replacing those Hamiltonian eigenstates with random states in their final expression.

Lev: The reliance on those specific assumptions is where I get cautious because in real hardware scenarios, we don't always have perfectly clean energy shells or guaranteed Haar-distributed initial conditions.

Kai: That’s a very important point, Lev; the authors themselves acknowledge that the third assumption about replacing eigenstates with random states is the most delicate part of their calculation.

Mira: They even corroborate this by noting that this replacement assumption can only be fully justified when restricting the analysis to an energy shell of eigenstates, which they confirm numerically using simulations of the one-dimensional Mixed Field Ising Model thirty-one <ref:2603.25387#pg1,when restricting the analysis to an energy shell of eigenstates>.

Lev: So, if we want to use this result in a real setting, we need to be very careful about where our system actually sits within that energy shell.

The paper's improvements: Kai: Looking at the improvements section of "A derivation of the late-time volume law for local operator entanglement," the authors are clarifying that they’ve derived an approximate expression for late-time LOE in terms of subsystem dimensions dA and d.

Mira: Specifically, in the limit where dA is small, but d is large, they arrive at a formula that looks like S two(rho(O)A(t)) about two (dA) -

one + (d − one) sigma squared (off-diag [sigma squared (diag - one): )] (D5) <ref:2603.25387#pg0>.

Lev: That formula gives us a concrete scaling relationship based on the dimensions of the subsystem and the system size, which is exactly what we need to predict how entanglement will behave if we scale up our physical device.

Kai: It provides a clear analytical handle on how that volume law manifests, moving past just observing linear growth numerically; it gives us a way to calculate those rates theoretically.

Mira: The improvement here is taking the theoretical structure—the Liouvillian basis and the averages—and applying the ETH assumption and then using Weingarten calculus to get this specific form involving sigma squared terms <ref:2603.25387#pg0>.

Lev: If we can use this, it suggests a way for us to benchmark our error correction codes against what we expect the entanglement entropy of a chaotic system to look like at late times.

Conclusion: Kai: So, wrapping up the discussion on "A derivation of the late-time volume law for local operator entanglement," it seems this paper provides a rigorous analytical path to understanding how LOE scales in chaotic many-body systems using established theoretical tools.

Mira: The main implication is that we now have a derived expression for the late-time LOE, contingent on those specific assumptions about non-resonance and ETH, which gives us a strong theoretical foundation connecting chaos indicators to observable scaling laws.

Lev: For my work in quantum error correction, the value here is that it provides a benchmark—a theoretical expectation—for how entanglement should evolve so we know what kind of noise or dynamics we are actually dealing with when designing codes.

Kai: I agree, Lev; having that analytical tool means our experimental validation isn't just checking if something scales linearly, but checking if it matches this specific derived function.

Mira: Overall, the paper contributes a concrete formula for late-time LOE scaling in chaotic systems while being very transparent about the assumptions they had to make to get there.

Lev: I think the explicit acknowledgment of when those assumptions are most justified, like on an energy shell, is what makes this result useful for connecting theory to what we can actually measure in a complex quantum environment.

Department of Physics, University of Helsinki · School of Physics, Trinity College Dublin

quant-ph, cond-mat.other

Submitted: 2026-03-26

Updated: 2026-10-07

Comments: 28 pages, 27 figures

Journal ref: J. Phys. A: Math. Theor. 59, 335301 (2026)

DOI: 10.1088/1751-8121/ae8f8a

Code: https://github.com/GICorrer/ETH_late_times_random_tensor_networks

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 83/100

The gist: Local Operator Entanglement (LOE) serves as an indicator of quantum chaos in many-body systems, and this paper provides an analytical derivation for its late-time behavior in chaotic systems that

Key concepts

Local Operator Entanglement (LOE)
LOE measures how entangled a specific local operator is within a many-body system. It is calculated using the 2-Rényi entropy of the vectorization of that operator, helping researchers distinguish between chaotic and integrable quantum behaviors.
Volume Law Scaling
This refers to a specific type of entanglement scaling where the entanglement grows proportionally to the number of local degrees of freedom. In chaotic systems, this means LOE scales with the size of the subsystem rather than growing indefinitely or saturating at a constant value.
Eigenstate Thermalization Hypothesis (ETH)
ETH is an assumption stating that for chaotic systems, matrix elements connecting different energy eigenstates are statistically random. This hypothesis allows researchers to simplify complex calculations by treating these matrix elements as random variables with specific statistical properties.
Liouvillian Basis
The Liouvillian describes how the state of a local operator evolves over time. The paper uses the eigenvectors of this operator's evolution (the Liouvillian basis) to simplify the time-dependent evolution into a manageable sum involving frequencies ($ ilde{ u}_m$) and matrix elements.

Terminology

Summary

Local Operator Entanglement (LOE) serves as an indicator of quantum chaos in many-body systems, and this paper provides an analytical derivation for its late-time behavior in chaotic systems that exhibits a volume-law scaling. This finding is significant because it offers a theoretical foundation for the linear growth and saturation observed numerically in chaotic dynamics, bridging the gap between numerical evidence and a complete analytical understanding of operator entanglement dynamics.

The Gist

An explicit formula displaying volume-law scaling for late-time LOE in chaotic systems is derived by expressing LOE in the Liouville eigenstate basis and relying on three main assumptions: a higherorder non-resonance condition for the Hamiltonian eigenenergies, the Eigenstate Thermalization Hypothesis (ETH) ansatz for the matrix elements of the initial local operator, and the replacement of Hamiltonian eigenstates with random states in the final expression for LOE.

Definition and Context of LOE

The paper defines Local Operator Entanglement (LOE) as the entanglement entropy of a vectorized local operator, defined as:

The LOE is defined as the entanglement 2−R´enyi entropy of the vectorization of O(t) given a bipartition of H [15]."

The quantity is used to distinguish chaotic and integrable behavior for short-range Hamiltonians. In chaotic systems, numerical studies show a characteristic linear increase with time, suggesting a volume law for entanglement at late times. The volume law refers to scaling proportional to the number of local degrees of freedom, or logarithmic in the Hilbert-space dimension of the subsystem.

Key Analytical Assumptions

The derivation relies on three central hypotheses:

  1. The “4 non-resonance condition” for the eigenvalues of the system Hamiltonian, which is assumed to hold for chaotic systems [27–29]. This condition ensures that certain sums of eigenenergies imply equality of eigenstate sets.

  2. The Eigenstate Thermalization Hypothesis (ETH) ansatz applied to expectation values of local operators, which is considered well-established in the chaotic regime [28–30]. The matrix elements are expressed as:

Oab = Omicroc(E)δab + e −S(E)/2 fO(E, ωab)∆ab,

where ∆ab are random real or complex numbers with zero mean and unit variance.

  1. The assumption that, for the purpose of computing the total LOE, the Hamiltonian eigenstates can be replaced by Haar-distributed random vectors, provided the system dimension is large. The authors acknowledge this assumption is most delicate and only fully justified when restricted to an energy shell or bulk of the spectrum [2].

Derivation via Liouvillian Basis and Averages

The computation proceeds by focusing on the long-time average of the exponential of LOE, defined as:

Tr [(ρ(O)A)2] = lim tmax→∞ 1/tmax ∫ tmax 0 dt Tr [(ρ(O)A(t))2].

The time evolution is analyzed using the Liouvillian L[⋅] and its eigenvectors, which form an effective basis of lower dimension, denoted by ωm⟩. The Heisenberg evolution is simplified to:

O(t)⟫ = K−1 ∑ m=0 e(iωmtNm ωm⟩ = K−1 ∑ m=0 e(iωmtNm Nm ωm⟩.

The long-time average purity, considering the 4 non-resonance condition, simplifies to three remaining terms involving traces of the ω matrices:

Tr[(ρ(O)A)2] = K−1 ∑ m=0 N 4 m TrA (ω(m)†ω(m)ω(m)†ω(m)) + K−1 ∑ m≠p=0 N 2 mN 2 p TrA (ω(m)ω(m† ωp ωp†)

  • K−1 ∑ m≠p=0 N 2 mN 2 p TrB (ω(m)†ω(m)ω(p† ωp)) (31).

Final Volume Law Expressions

By applying the ETH assumption and replacing eigenstates with Haar random states, the expression is transformed into Equation (40), which can be computed analytically using Weingarten calculus. This leads to an approximated expression for late-time LOE in terms of subsystem dimensions dA and d:

**In the limit dA = O(1), d ≫ 1, the formula obtained is approximately: ⟨S2(ρ(O)A(t))⟩ ≈ 2 ln(dA) − ln [1 + (d − 1)σ squared off-diag [σ squared diag − 1]] (D5).

Improvements for AI systems

As a fastidious and diligent researcher, I have analyzed this paper's contribution to understanding the late-time behavior of Local Operator Entanglement (LOE) in chaotic many-body systems. The core finding is the derivation of an analytical formula for LOE scaling with a volume law, contingent on specific assumptions (4 non-resonance condition, ETH ansatz, and replacement of eigenstates with random states).

Based on this scientific framework, here are the specific improvements that can be made to AI systems:

  1. Improvement in Quantum Chaos Characterization and Modeling:

  2. Improvement in Entanglement Dynamics Prediction for Complex Systems:

  3. Improvement in Machine Learning for Hamiltonian/Operator Estimation:

Here is a detailed breakdown of what the improved AI system can achieve in each area:

  1. AI systems can perform highly accurate, analytical predictions of late-time entanglement scaling (volume law) for quantum chaotic many-body systems, moving beyond purely numerical simulation results. This allows for the rapid characterization of whether a given physical system exhibits chaotic dynamics based on its expected LOE growth rate.

  2. The AI can be used to predict the entanglement dynamics in complex, large-scale quantum simulations (e.g., simulating quantum field theories or condensed matter models) with high fidelity, providing a benchmark for validating the assumptions of ETH and non-resonance conditions in those specific regimes.

  3. The AI can be trained to estimate the matrix elements of local operators within the eigenbasis using sophisticated tensor network methods (as suggested by Appendix D), leading to more efficient and accurate calculation of entanglement measures like LOE in high-dimensional Hilbert spaces where exact diagonalization is infeasible.

Sources

Related papers