Provable quantum thermalization without statistical averages

Abstract: We develop a rigorous system-agnostic method to predict quantum thermalization in an overwhelming fraction of accessible pure states in a many-body system, entirely in terms of certain out-of-time-ordered correlators of few-body observables. In contrast to previous rigorous results on thermalization with semiclassical counterparts, our method is not limited to statistical averages of observables, such as time averages in ergodicity or state averages in mixing. Moreover, consistent with such approaches, we retain the advantage of not requiring a detailed knowledge of energy eigenstate structure or thermodynamically large times, which can become intractable for systems with more than a handful of particles. Our approach is centered on a geometric result that connects thermalization to the alignment of high dimensional subspaces in a Hilbert space, which is determined by the saturation of "controllably nonlocal" out-of-time-ordered correlators. This formalism reduces the problem of establishing pure state quantum thermalization at finite times in almost all complex many-body states to a theoretically or experimentally accessible study of few-body correlators, even in thermodynamically large systems.

• The paper demonstrates that pure-state thermalization can be proven without statistical averages by using out-of-time-ordered correlators (OTOCs) and finite-time dynamics.
• It introduces a rigorous framework based on subspace geometry and variance analysis of OTOCs to establish necessary and sufficient conditions for thermalization.
• Numerical results validate that controlled nonlocal OTOC measurements provide practical benchmarks for pure state thermalization without requiring long-time extrapolation.

Provable Quantum Thermalization Without Statistical Averages: An Expert Summary

Motivation and Problem Statement

The paper "Provable quantum thermalization without statistical averages" (2604.02417) addresses a fundamental gap in quantum statistical mechanics: the ability to rigorously diagnose thermalization in individual pure states of a many-body quantum system without invoking statistical averages, such as time or state averaging. Classical and conventional quantum statistical mechanics rely on ergodic or mixing properties, which inherently involve such averages and are strictly relevant only for infinite-time (thermodynamically large time) behavior. Moreover, typicality arguments are mathematically insensitive to physically relevant basis states, since Haar measure assigns measure zero to the sets of states often encountered in practical settings (e.g., computational basis states).

The central challenge outlined is to extrapolate from finite, experimentally or computationally accessible information—specifically, few-body dynamics over finite times—to robust predictions of thermalization within an overwhelming fraction of pure states, irrespective of not knowing energy eigenstates or requiring infinite time. Previous rigorous approaches, including the eigenstate thermalization hypothesis (ETH) and recently developed autocorrelation-based methods, invariably require statistical averaging and thus fall short of delivering results for single pure states at specific times.

Main Results and Analytical Framework

Elimination of Statistical Averages via OTOCs

The paper's core innovation is the establishment of a necessary and sufficient criterion for quantum thermalization in almost all basis states (within a large subspace), using out-of-time-ordered correlators (OTOCs) of few-body observables. In particular, it is shown that the instantaneous alignment of high-dimensional subspaces in Hilbert space—controlled by the factorization property of certain OTOCs—is directly linked to thermalization in pure states without averages.

Formally, for a projector observable $R$ supported on $N_S$ qubits, and a core state projector $\psi$ supported on $N_\sigma$ qubits (with $N_\sigma \gg N_S$), the second- and fourth-order correlators

$$G_{R\psi}^{(2)}(t) = \frac{D_\sigma}{D} \text{Tr}[\psi(t) R], \qquad G_{R\psi}^{(4)}(t) = \frac{D_\sigma}{D} \text{Tr}[R \psi(t) R \psi(t)]$$

map thermalization of $R$ in almost all bath basis states, for fixed $\psi$, at time $t$, to the near-factorization $G_{R\psi}^{(4)}(t) \approx [G_{R\psi}^{(2)}(t)]^2$. This "controllably nonlocal" OTOC (with extended support in the core) must decay to sufficiently small values for thermalization without averages to hold to a prescribed resolution.

Subspace Geometry and Thermalization

The analysis is grounded in the geometric properties of operator subspaces, leveraging principal angle distributions between projectors. If the variance $N_S$0 of the cosines squared of principal angles (as encoded by $N_S$1 and $N_S$2) is small, then quantum thermalization is provable to the corresponding accuracy for most pure states in any orthonormal basis of the bath. This result is captured by rigorous bounds (see Theorem 1 and Corollary statements) and is both necessary and sufficient: if instantaneous pure-state thermalization occurs for all bases, then the OTOC-based variance must vanish correspondingly.

Typicality and Scalability

Typicality arguments and concentration-of-measure results on the unitary group are invoked to quantify fluctuations of these correlators, demonstrating that for 'controllably nonlocal' OTOCs (with $N_S$3), the requisite small variance is generic within high-dimensional Hilbert spaces. The results scale favorably in the thermodynamic limit, provided the core system remains finite and larger than the observed subsystem—a condition explicit in the cubic scaling dependence derived for experimental feasibility.

Numerical Results and Strong Claims

• Quantitative bounds: For single-qubit thermalization to $N_S$4 accuracy in $N_S$5 of bath basis states, nonlocal OTOCs with $N_S$6 must be measured to $N_S$7 accuracy; for $N_S$8 of bath states and $N_S$9 accuracy, $\psi$0 and $\psi$1 resolution suffice.
• Necessity and sufficiency: The smallness of the OTOC variance is not merely sufficient but necessary for pure state thermalization across all bases.
• No classical analog: The established criteria are sensitive to operator ordering, failing in the classical limit—thus they capture intrinsically quantum phenomena inaccessible via classical or semiclassical theory.
• No extrapolation in time: Unlike autocorrelator-based methods, the OTOC condition allows no extrapolation from finite-time observations to infinite-time forecasts when statistical averages are absent.

These strong, non-asymptotic numerical requirements make rigorous experimental diagnosis of quantum thermalization in pure states—without averages or typicality assumptions—a demanding but achievable goal given advancing measurement and computational capabilities.

Practical and Theoretical Implications

• Experimental diagnosis: The scalability of OTOC-based measurements via control qubits is discussed, emphasizing finite experimental resources and the exponential suppression of accessible purity protocols (randomized measurements).
• Hierarchy of timescales: OTOC saturation identifies the slowest collective thermalization across bases, while autocorrelator decay describes typical rates; the spectral form factor bounds the fastest averaged rates.
• Quantum statistical mechanics framework: The results advance quantum statistical mechanics towards a formalism in which thermalization can be predicted from finite-time, few-body correlator measurements, obviating reliance on inaccessible energy eigenstates or statistical averages, at least for instantaneously accessible times.
• Future developments: The paper points to the need for new techniques to extrapolate OTOC behavior to long times, or at least classify system families where such extrapolation is feasible. It also raises the question of intermediates between typicality and universal pure-state thermalization in all bases.

Conclusion

This work rigorously ties the problem of quantum thermalization in pure states, without resorting to statistical averages, to the instantaneous saturation of controllably nonlocal out-of-time-ordered correlators. The necessity and sufficiency of OTOC-based variance smallness constitute a robust criterion that is amenable to theoretical computation and scalable experimental measurement. While predictions are restricted to times of observation, and extrapolation to infinite times remains elusive, the framework surpasses classical approaches and prior quantum results anchored to statistical averages. The implications for both foundational understanding and practical experiments in quantum statistical mechanics are substantial, delineating a clear path for future exploration of pure-state thermalization diagnostics, model-independent forecasting, and complexity theory in quantum many-body systems.