Quantum Ergodicity Threshold (QET)

• Quantum Ergodicity Threshold (QET) is a sharp criterion that defines the transition from integrable or localized behavior to universal, random-matrix ergodicity in quantum systems.
• It employs quantitative measures such as Fisher information, quantum variance, and frame potentials to demonstrate the exponential suppression of temporal fluctuations and approach to equilibrium.
• QET provides clear operational diagnostics across various models—including many-body, driven, and molecular systems—enabling a unified understanding of thermalization and ergodic transitions.

The Quantum Ergodicity Threshold (QET) is a sharply defined criterion, observable, or coupling scale at which a quantum system transitions from non-ergodic to ergodic dynamics. This concept is critical in quantum statistical mechanics, quantum information theory, and the study of thermalization, providing precise, physically meaningful ways to detect when a system's eigenstates, dynamics, or observables exhibit universal, random-matrix-like ergodicity as opposed to integrable or localized behavior.

1. Fundamental Definition and Mathematical Characterization

In closed quantum systems with Hilbert space dimension $D$, the QET is formulated as the point where a suitably chosen quantitative measure attains a value signifying that all initial-state dependent or observable-specific deviations from equilibrium (or randomness) vanish up to corrections that decay exponentially with system size. In the framework of global ergodicity for isolated systems, the QET is marked by the quantum Fisher information $F(A, H)$ of an observable $A$ under a non-degenerate Hamiltonian $H$ reaching its minimal (Cramér–Rao) bound: $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$ where $S = \ln D$ is the Hilbert-space entropy. At or below this threshold, time-averages of $A$ in any generic initial state converge, up to exponentially small corrections, to the microcanonical average. Temporal fluctuations become negligible, off-diagonal matrix elements are maximally suppressed (stronger than ETH scaling), and operator complexity saturates chaotic bounds (Gomez, 2019).

Analogously, for quantum graphs, driven systems, or many-body models, the QET is specified by the vanishing of quantum variance, the saturation of frame potentials, or the robust scaling of eigenstate or local observable diagnostics. In all cases, the QET demarcates the transition between integrable/localized phases and universal, random dynamics.

2. Microscopic Criteria and Observable Thresholds

Fisher Information Ergodicity Criterion

The Fisher-information-based criterion is both necessary and sufficient in chaotic, nondegenerate settings. Given

$$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$

where the overlines denote infinite-time-averaging, ergodicity is certified when

$F(A, H) \leq F_Q^{\rm (crit)} = D^{-2}~.$

Saturation of the lower (Cramér–Rao) bound $e^{-2S}$ implies full quantum ergodicity—time-fluctuations are $F(A, H)$0, off-diagonal $F(A, H)$1 are exponentially suppressed, and operator complexity/time scales such as the Rindler-time and circuit depth scale as $F(A, H)$2 (Gomez, 2019).

Graph Variance Bound and Cycle Structure

For quantum graphs with equi-transmitting boundary conditions, the QET can be expressed in terms of the decay of the quantum variance $F(A, H)$3 of observables $F(A, H)$4: $F(A, H)$5 where $F(A, H)$6 counts directed bonds on cycles of length $F(A, H)$7, and $F(A, H)$8 is the spectral gap of the adjacency matrix. The QET is the minimal $F(A, H)$9 such that both error terms vanish in the large-$A$0 limit; typically, $A$1, with cycle density subexponential in $A$2 (Brammall et al., 2015).

Frame Potentials and Dynamical Designs

In periodically or quasiperiodically driven (Floquet) systems, quantum ergodicity is tied to the dynamical generation of exact or approximate unitary designs, governed by the $A$3th frame potential. Complete unitary ergodicity requires

$A$4

independent drive frequencies (for dimension $A$5), ensuring the temporal ensemble is indistinguishable from Haar measure in all moments. For single-frequency drives, only finite-$A$6 unitary ergodicity is achievable; the threshold period $A$7 is required to mimic Haar statistics to the $A$8th moment (Pilatowsky-Cameo et al., 2024).

3. Model-Dependent QETs: Many-Body, Molecular, and Graphical Systems

Spin Chains and Many-Body Hamiltonians

In many-body spin chains, the QET corresponds to the value of an external parameter (e.g., interaction strength $A$9) where the scaling of ergodicity-sensitive metrics (e.g., the variance gap $H$0 or inverse spread $H$1 of local operator covariance matrices) switches from polynomial to exponential with system size $H$2:

• Integrable: $H$3 polynomially in $H$4
• Ergodic: $H$5 exponentially in $H$6

This sharply discriminates between ergodic (ETH-satisfying) and integrable regimes (Gjonbalaj et al., 4 Mar 2026).

Spectral and Eigenstate-Based Landmarks

Alternative QET diagnostics employ IR entropy/diversity, inverse participation ratio (IPR), or the $H$7-parameter of level spacings. The QET $H$8 is operationally identified as the $H$9 where the ensemble-averaged IR diversity $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$0 reaches the random-matrix value (GOE: $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$1 for real systems), or where mean $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$2 crosses midway between Poisson ($F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$3) and GOE ($F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$4) (Ho et al., 2017).

Ergodicity in Coupled Chaotic Subsystems

In bipartite Bose–Hubbard models, the QET ($F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$5) marks the onset of ETH: the critical coupling $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$6 where eigenstates fill the "energy shell" ergodically and participation ratios match GOE statistics. This generally occurs at dimensionless coupling $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$7, with $F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$8 the energy-shell width in units of the mean level spacing (Vardi et al., 31 Jul 2025). Notably, this eigenstate QET is distinct from the weaker threshold for level-repulsion/spectral chaos.

Molecular and Rotational Degrees of Freedom

In molecular systems, such as SCCl$F_Q^{\mathrm{(crit)}} = D^{-2} = e^{-2S}$9, the QET is manifest as a crossover in IPR and entropy measures for highly excited eigenstates, indicating partial delocalization but never full ergodicity even near dissociation threshold—a reflection of complex phase-space structure and non-random connectivity (Keshavamurthy, 2013).

For icosahedral C$S = \ln D$0, a sequence of rotational QETs is observed as angular momentum increases, driven by competition among high-rank symmetric tensor invariants. Here, ergodicity-breaking transitions occur at angular momenta $S = \ln D$1 far below the vibrational IVR threshold, controlled by symmetry, tensor structure, and avoided crossings with nearby vibrational states (Liu et al., 2023).

4. Multiscale and System-Size Effects in Quantum Ergodicity

The QET may depend on spatial or subsystem scale. In 2D disordered Floquet Heisenberg systems, ergodicity onset occurs hierarchically: small spatial patches (e.g., $S = \ln D$2) reach random-matrix-theory values of marginal collision entropy at lower coupling $S = \ln D$3 than larger patches ($S = \ln D$4, $S = \ln D$5), with $S = \ln D$6. There is no single sharp global QET, but rather a family of thresholds $S = \ln D$7 parameterized by patch size, accessible via direct measurement of collision entropy or Rényi-2 entropy on quantum hardware (Alam et al., 12 Mar 2026).

Finite-size scaling and classical/quantum simulation boundaries can also affect how sharply and at what values the QET is observable, but the convergence of patchwise ergodicity indicators for increasing total system size supports the robustness of operational QETs even in near-term devices.

5. Comparative Analysis with ETH, Random Matrix Theory, and Classical Chaos

QET criteria summarized above are stronger than traditional ETH or RMT-linked notions of chaos:

• ETH typically demands $S = \ln D$8; QET demands summing these over all off-diagonal elements suppresses fluctuations to $S = \ln D$9 (Gomez, 2019).
• RMT threshold for spectral chaos (level-repulsion) is generically lower than the QET for full ergodicity of eigenstates and observables (Vardi et al., 31 Jul 2025).
• In classical scenarios, QET analogues relate to percolation thresholds in connectivity graphs, mixing-to-chaos crossovers, and the covering of phase-space by classical trajectories (Brammall et al., 2015, Ho et al., 2017).

QET therefore provides a sharp, observable-independent, and often system-size-independent demarcation between regimes where random-matrix theory, ETH, and Haar statistics apply, and those where integrability, localization, or symmetry constraints dominate.

6. Physical Consequences and Implications

Thermalization and Complexity

At or beyond the QET, quantum systems thermalize rapidly: observable fluctuations are exponentially suppressed, operator growth and complexity reach their maximal bounds, and equilibration timescales match those predicted by information-theoretic and black-hole models (e.g., scrambling timescales $A$0 or macroscopic decoherence times $A$1 for $A$2 macro-sectors) (Gomez, 2019).

Diagnostic Hierarchy

The QET unifies and extends a wide array of diagnostic tools:

• Fisher information metrics and variance gaps for operator/observable dynamics (Gomez, 2019, Gjonbalaj et al., 4 Mar 2026)
• Participation ratios and information-theoretic entropy for eigenstates (Keshavamurthy, 2013, Ho et al., 2017)
• Frame potentials and $A$3-design conditions for driven and open-system dynamics (Pilatowsky-Cameo et al., 2024)
• Spectral statistics and level-spacing ratios for universal transitions (Ho et al., 2017)

Model-Specific Nuances

While QET is universal in its mathematical formulation, the exact threshold and the sharpness of the transition depend on system connectivity, underlying symmetries, and the degree of randomness of the couplings. Models with inhomogeneous coupling, nontrivial conservation laws, or special point-group symmetries exhibit multi-stage or broad crossovers rather than sharp transitions.

Model Class	QET Observable/Parameter	Threshold Scaling
Global nondegenerate system	$A$4 (Fisher information)	$A$5
Quantum graphs (expanders)	Variance $A$6	$A$7, cycle density $A$8
Driven (Floquet/quasi-periodic)	No. of drive frequencies $A$9	$$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$0
Coupled chaotic subsystems	Inter-coupling $$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$1	$$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$2
Disordered/interacting many-body	Operator variance gap $$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$3	Polynomial $$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$4 exponential scaling
Molecular rotational modes (C$$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$5)	Angular momentum $$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$6	$$G(A, |\psi\rangle) \equiv \overline{\langle A(t)\rangle^2} - (\overline{\langle A(t)\rangle})^2 \leq F(A, H)~,$$7 by tensor invariant crossover

7. Summary and Theoretical Synthesis

The Quantum Ergodicity Threshold provides a precise, physically invariant criterion for the transition to universal, thermal, and random dynamics in quantum systems. Its formulations—via Fisher information, quantum variance, frame potentials, entropy measures, or participation ratios—quantify the statistical complexity and observable-universality required for the applicability of statistical mechanics, quantum information protocols, and random-matrix theory. QET subsumes and sharpens the eigenstate thermalization hypothesis, random-matrix universality criteria, and dynamical design thresholds, offering unified scaling laws and operational diagnostics in a diversity of quantum models and experimental contexts (Gomez, 2019, Brammall et al., 2015, Pilatowsky-Cameo et al., 2024, Gjonbalaj et al., 4 Mar 2026, Keshavamurthy, 2013, Vardi et al., 31 Jul 2025, Liu et al., 2023, Ho et al., 2017, Alam et al., 12 Mar 2026).