Out-of-time-order correlator computation based on discrete truncated Wigner approximation

Tatsuhiko Shirai, Takashi Mori · 2025

Read Out-of-time-order correlator computation based on discrete truncated Wigner approximation by Tatsuhiko Shirai, Takashi Mori (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-14221.txt.

Introduction

Tatsuhiko Shirai and Takashi Mori compute chaos probes in quantum spin systems with an approximation method, in a 2025 paper. The probe, an out-of-time-order correlator, tracks how information scrambles. Exact computation is costly in big systems. Their method adapts a discrete phase-space technique. The systems studied interact across long range. Coupling fades with distance as a power law. A squared commutator of local operators serves as the measure.

The method matched exact dynamics in strongly long-range cases. It captured the average spread of information at all times. In weakly long-range cases it missed details. The fastest scrambling front escaped it. The team used it on system-size scaling. Scrambling time scaling changed character with interaction range. That split was qualitative, not just numeric. The work gives simulators a new tool for scrambling studies. Its limits are mapped as carefully as its wins.

Catalog record

CREATORTatsuhiko Shirai, Takashi Mori
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-14221

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Out-of-time-order correlator computation based on discrete truncated Wigner approximation Tatsuhiko Shirai, Takashi Mori Abstract We propose a method based on the discrete truncated Wigner approximation (DTWA) for computing out-of-time-order correlators. This method is applied to long-range interacting quantum spin systems where the interactions decay as a power law with distance. As a demonstration, we use a squared commutator of local operators and its higher-order extensions that describe quantum information scrambling under Hamilton dynamics. Our results reveal that the DTWA method accurately reproduces the exact dynamics of the average spreading of quantum information (i.e., the squared commutator) across all time regimes in strongly long-range interacting systems. We also identify limitations in the DTWA method when capturing dynamics in weakly long-range interacting systems and the fastest spreading of quantum information. Then we apply the DTWA method to investigate the system-size dependence of the scrambling time in strongly long-range interacting systems. We reveal that the scaling behavior of the scrambling time for large system sizes qualitatively changes depending on the interaction range. This work provides and demonstrates a new technique to study scrambling dynamics in long-range interacting quantum spin systems. I Introduction Out-of-time-order correlator (OTOC) [ 1 ] has attracted attention in non-equilibrium statistical mechanics, quantum information, and quantum gravity [ 2 ] . The time evolution of the OTOC estimates the scrambling time t_{*} when local perturbation propagates to the entire system. Systems that exhibit logarithmic scrambling time, t_{*}\sim\beta\hbar\log N , where \beta is the inverse temperature, \hbar is the Planck constant, and N is the number of degrees of freedom, are referred to as fast scramblers [ 3 , 4 ] , with a holographic duality to black hole being explored [ 5 ] . Meanwhile, ballistic information propagation in a chaotic spin system with short-range interaction [ 6 ] implies a polynomial scrambling time, t_{*}\sim N^{1/d} , where d is the space dimension. Coherently simulating a reverse time evolution has facilitated the measurements of the OTOC in trapped ions [ 7 , 8 ] , nuclear magnetic resonance systems [ 9 ] , and superconducting circuits [ 10 ] . Long-range interacting quantum spin systems, where the interaction decays as a power law J\sim r^{-\alpha} with distance r , exhibit intriguing dynamical properties [ 11 ] . Rigorous bounds on the scrambling time indicate that translationally invariant spin chains with extensive energy are not fast scramblers, i.e. t_{*}\sim N^{\gamma} [ 12 , 13 , 14 ] . Although the scrambling dynamics were numerically investigated in previous works [ 15 , 16 , 17 ] , accurately estimating the exponent \gamma is challenging due to the exponentially growing Hilbert space and the strong finite-size effects. Therefore, developing approximated methods to tackle this problem is crucial. Existing approximated methods are insufficient to characterize scrambling dynamics. A classical approach (as defined in Sec. II.3 ) completely neglects the effects of quantum fluctuations on temporal correlation functions. Truncated Wigner approximation [ 18 ] , which relies on a continuous Wigner function defined over a continuous phase space, fails to capture the spatial structure of correlations. Since scrambling dynamics is intrinsically tied to the propagation of local correlations throughout the system, a method capable of resolving the dynamics at the level of individual spins is necessary. One promising approach to address these problems is the discrete truncated Wigner approximation (DTWA) [ 19 ] . DTWA offers a semiclassical method that enables efficient simulation of quantum dynamics by representing quantum states in a discrete phase space [ 20 ] . DTWA has been shown to accurately reproduce collective observables, spatial correlation functions, and relative entropy in the long-range interacting systems [ 19 , 21 , 22 ] . However, DTWA-based approaches for computing time-correlation functions have not yet been developed. In this study, we propose a DTWA-based method for computing temporal correlation functions including OTOCs. The method is applied to calculate a squared commutator of local operators, along with its higher-order extensions, to describe quantum information spreading under the Hamilton dynamics. We benchmark the DTWA method across systems with varying values of \alpha , covering from weakly (\alpha>d) to strongly (\alpha\leq d) long-range interactions. In the strongly long-range interacting regime, unlike the weakly long-range case, the system exhibits nonadditivity in its energy. Our numerical results reproduce the exact dynamics for the average spreading of quantum information (i.e., squared commutators) across all time regimes in strongly long-range interacting systems. Furthermore, we identify the limitations of the DTWA method when simulating scrambling dynamics in weakly long-range interacting systems and the fastest spreading of quantum information. In addition, we observe that the approximated dynamics of an autocorrelation function holds valid over short-time scales, with these timescales being shorter as \alpha increases. These findings clarify the applicability range of the DTWA method in exploring scrambling dynamics in long-range interacting quantum spin systems. Finally, we apply the DTWA method to investigate the system-size dependence of the scrambling time in strongly long-range interacting regime and compare the numerical result with a theoretical lower bound. We reveal that the scaling behavior of the scrambling time for large system sizes qualitatively changes depending on the value of \alpha . This paper is organized as follows. In Sec. II , we present a model for long-range interacting systems, the DTWA method for computing time-correlation functions including OTOCs, and its benchmark results. In Sec. III , we provide a theoretical bound and the numerical results using the DTWA approach for the scrambling time. Section IV summarizes this paper with some future directions. Appendix provides the derivation of the DTWA expression for OTOCs, efficient exact simulation method at \alpha=0 , and system-size dependences of the DTWA method. II Model and Methods II.1 Long-range interacting quantum spin systems We consider a quantum spin system on a lattice \Lambda\in\{1,\ldots,N\} . The Hamiltonian is given by \hat{H}=\sum_{\begin{subarray}{c}i,j\in\Lambda\\ (j>i)\end{subarray}}\hat{\bm{\sigma}}_{i}\bm{J}_{ij}\hat{\bm{\sigma}}_{j}^{\intercal}+\sum_{i\in\Lambda}\bm{h}\hat{\bm{\sigma}}_{i}^{\intercal}, (1) where \bm{\sigma}_{i}=(\hat{\sigma}_{i}^{x},\hat{\sigma}_{i}^{y},\hat{\sigma}_{i}^{z}) is the vector for Pauli spin operators acting on site i\in\Lambda and \intercal denotes the transpose. \bm{J}_{ij}=\{J_{ij}^{ab}\}_{a,b\in\{x,y,z\}} represents the interaction matrix between spins i and j , with J_{ij}^{ab}=J_{ji}^{ba} , and \bm{h}=(h^{x},h^{y},h^{z}) describes a uniform magnetic field, respectively. The interaction strength decays as a power of \alpha with distance r_{ij} , given as J_{ij}^{ab}=\frac{J^{ab}}{\mathcal{N}(\alpha)}r_{ij}^{-\alpha}. (2) Although the proposed DTWA method is applicable regardless of boundary conditions and lattice topologies, we herein assume a one-dimensional lattice with a periodic boundary condition. Then r_{ij}=\min\{|i-j|,N-|i-j|\} . The interaction is called strongly long range when 0\leq\alpha\leq 1 , whereas weakly long range when \alpha>1 [ 11 ] . The normalization of \mathcal{N}(\alpha)=\sum_{i=2}^{N}r_{1i}^{-\alpha} is known as the Kac prescription [ 23 ] so that the energy per spin is finite in the thermodynamic limit even at 0\leq\alpha\leq 1 . The model is reduced to an infinite-range model at \alpha=0 , whereas a short-range model with nearest-neighbor interaction at \alpha\to\infty . In numerical simulations, we adopt J^{ab}=\delta_{az}\delta_{bz} and (h^{x},h^{y},h^{z})=(0.9045,0,0.809) , where the eigenstate thermalization hypothesis was numerically shown in the limit of \alpha\to\infty [ 24 ] . Here, \delta_{ab} is the Kronecker’s delta. II.2 DTWA method to the OTOC We briefly review the DTWA method for computing \sigma_{i}^{a}(t)=\mathrm{Tr}(\hat{\sigma}_{i}^{a}(t)\rho)\eqqcolon\langle\hat{\sigma}_{i}^{a}(t)\rangle [ 19 ] , where i\in\Lambda , a\in\{x,y,z\} , \hat{\sigma}_{i}^{a}(t)=e^{\mathrm{i}\hat{H}t}\hat{\sigma}_{i}^{a}e^{-\mathrm{i}\hat{H}t} , and \rho is the density matrix of the system. Here we take \hbar=1 . Let us

Read the complete text