Universal Growth of Krylov Complexity Across A Quantum Phase Transition

Basic Information

• Paper ID: 2510.13947
• Title: Universal Growth of Krylov Complexity Across A Quantum Phase Transition
• Authors: András Grabarits, Adolfo del Campo
• Classification: quant-ph
• Publication Date: October 17, 2025
• Paper Link: https://arxiv.org/abs/2510.13947

Abstract

This paper investigates the statistical properties of diffusion complexity in Krylov space for quantum systems undergoing quantum phase transitions. Using non-adiabatic Magnus expansion, the evolution is mapped to an effective one-dimensional hopping model. For the transverse-field Ising model, an exact connection is established between complexity growth and Kibble-Zurek defect scaling: all cumulants of complexity exhibit the same power-law scaling as defect density, with coefficients matching the mean value, and the complete distribution asymptotically approaches a Gaussian distribution. These results provide general scaling arguments for complexity growth in arbitrary second-order quantum phase transitions.

Research Background and Motivation

1. Problems to be Addressed:
   • Complexity growth in quantum many-body dynamics is fundamental to thermalization, chaos, and the computational difficulty of describing quantum evolution
   • Characterization of operator growth and Krylov complexity in time-dependent settings remains largely unexplored
   • Need to determine whether physical quantities such as complexity also exhibit non-equilibrium universality
2. Problem Significance:
   • Krylov subspace methods have become a universal framework for quantifying complexity growth
   • Possess practical value in quantum control and variational quantum algorithms
   • Connect complexity growth to universal features of critical quantum dynamics
3. Limitations of Existing Approaches:
   • Existing general frameworks for time-dependent settings employ Floquet operators with non-local time, difficult to apply in many-body systems
   • Lack of theoretical framework connecting complexity growth to quantum phase transition universality
4. Research Motivation:
   • Develop exact and universal analytical frameworks applicable to time-dependent settings
   • Establish connection between Krylov complexity and Kibble-Zurek mechanism
   • Provide universal framework for complexity growth near critical points

Core Contributions

1. Proposed a new Krylov subspace method based on non-adiabatic Magnus operators, applicable to time-dependent settings of driven quantum systems
2. Established exact connection between complexity growth and Kibble-Zurek defect scaling, proving that all complexity cumulants in the transverse-field Ising model follow the same power-law scaling
3. Demonstrated that Krylov complexity statistics converge to universal Gaussian distribution within the KZ scaling regime, with complete distribution exhibiting universal behavior
4. Provided general scaling arguments applicable to arbitrary second-order quantum phase transitions, generalizing results to broader categories of critical systems

Methodology Details

Task Definition

Investigate the statistical properties of Krylov complexity for quantum systems driven at finite rates across quantum phase transitions, establishing universal connections between complexity growth and phase transition dynamics.

Model Architecture

1. Non-adiabatic Magnus Operator Construction

For driven systems $H(t) = \sum_n E_n(t)|n(t)\rangle\langle n(t)|$, define the non-adiabatic time evolution operator: $U(t) = U(t)U_{ad}(t)^\dagger$

where $U_{ad} = \sum_n |n(t)\rangle\langle n(0)|$ is the parallel transport operator. The non-adiabatic Magnus operator is defined as: $\Omega(t) = i\log(U(t))$

The time-evolved state can be expressed as: $|\psi(t)\rangle = e^{-i\Omega(t)}|GS(t)\rangle = \sum_{m=0}^{\infty} \frac{(-i)^m \Omega^m(t)}{m!}|GS(t)\rangle$

2. Krylov Basis Construction

The corresponding Lanczos algorithm is: $|K_{n+1,t}\rangle b_{n+1,t} = \Omega(t)|K_{n,t}\rangle - a_{n,t}|K_{n,t}\rangle - b_{n,t}|K_{n-1,t}\rangle$

where:

• $a_{n,t} = \langle K_{n,t}|\Omega(t)|K_{n,t}\rangle$
• $b_{n,t} = \langle K_{n-1,t}|\Omega(t)|K_{n,t}\rangle$

3. Transverse-Field Ising Model Analysis

TFIM Hamiltonian: $\hat{H}(t) = -J\sum_{j=1}^L [\sigma_j^z \sigma_{j+1}^z + g(t)\sigma_j^x]$

In fermionic representation, each momentum mode corresponds to an independent two-level system: $\hat{H} = 2\sum_k \hat{\psi}_k^\dagger H_k(t) \hat{\psi}_k$

where $H_k(t) = (g(t) - \cos k)\tau_z + \sin k \tau_x$.

Technical Innovations

1. Time Locality: Krylov bases and Lanczos coefficients at different times are generated independently, avoiding time non-locality issues
2. Exact Analytical Results: Obtain exact expressions for Lanczos coefficients and Krylov wavefunctions within the KZ scaling regime: $b_n \sim L^{1/2}\tau^{-1/4}\sqrt{n}, \quad a_n \sim L\tau^{-1/2}$
3. Poisson Statistics: Krylov wavefunctions exhibit Poisson distribution at leading order approximation: $|\phi_n(\tau)|^2 \sim \frac{(2CL\tau^{-1/2})^n e^{-2CL\tau^{-1/2}}}{n!}$

Experimental Setup

Dataset

• Transverse-field Ising model with system sizes $L = 200, 400$
• Effective system sizes $L_{eff} = 46, 48$ (for Lanczos algorithm)
• Driving time range: $\tau = 2$ to $\tau = 750$

Evaluation Metrics

• First three cumulants of Krylov complexity: $K_1$ (mean), $K_2$ (variance), $K_3$ (third central moment)
• Quality of Gaussian fit to complexity distribution
• Conformance to KZ scaling $\tau^{-1/2}$

Implementation Details

• Linear driving protocol: $g(t) = t/\tau$
• Initialization in paramagnetic phase at $t = -g_0\tau$
• Termination at $g(0) = 0$
• Computational complexity limits exact numerical simulations to $L \approx 48$

Experimental Results

Main Results

1. Universal Scaling of Cumulants: All complexity cumulants follow the same power law within the KZ scaling regime: $K_q \approx 2CL\tau^{-1/2}$
   This is in complete agreement with the KZ scaling of defect density $n \sim \tau^{-1/2}$.
2. Gaussian Distribution Convergence: Within the KZ scaling regime $L\tau^{-1/2} \gg 1$, the complexity distribution converges to a Gaussian: $P_K(n) \approx \frac{\exp[-(n-K_1)^2/(2K_2)]}{\sqrt{2\pi K_2}}$
3. Lanczos Coefficient Scaling: Numerical results verify the theoretically predicted scaling behavior, with data from all different $\tau$ and $L$ collapsing onto a single curve after appropriate rescaling.

Time Evolution Analysis

• Complexity exhibits sharp growth near the critical point, capturing the interplay between quantum phase transition and complexity growth
• Shows non-universal oscillatory behavior away from the critical point
• Gradually converges to asymptotic values within the symmetry-broken phase

Scaling Collapse Verification

Figure 1 displays off-diagonal Lanczos coefficients for different system sizes and driving times, which perfectly collapse onto a single curve after appropriate rescaling by $\tau^{1/4}L^{-1/2}$.

Related Work

Krylov Complexity Theory

• Universal operator growth hypothesis proposed by Parker et al.
• Applications of Krylov subspace methods in quantum chaos and entanglement growth
• Complexity studies in time-independent settings

Kibble-Zurek Mechanism

• Universal framework for defect formation in second-order quantum phase transitions
• Verification of KZ scaling in various quantum systems
• Universality of defect statistics

Quantum Critical Dynamics

• Adiabatic-impulse approximation theory
• Landau-Zener transitions in many-body systems
• Universality in non-equilibrium quantum phase transitions

Conclusions and Discussion

Main Conclusions

1. Established universal connection between complexity and phase transition dynamics: All cumulants of Krylov complexity follow the same KZ scaling as defect density
2. Demonstrated universality of complexity statistics: Within the KZ regime, complexity distribution converges to universal Gaussian form
3. Provided general framework: Results generalize to arbitrary free fermionic systems in $d$ dimensions with $(d-D)$-dimensional critical manifolds

Generalized Results

For general critical systems, complexity cumulants follow: $K_q \sim 2CL^{d-D}\tau^{-\alpha(d-D)}$

where $\alpha$ and $d-D$ characterize the excitation probability and dimensionality of the critical manifold, respectively.

Limitations

1. Computational Complexity Constraints: The direct sum form of the Magnus operator leads to exponentially growing computational cost, limiting exact numerical simulations to relatively small system sizes
2. Effective Range of Approximations: Leading-order approximation is valid only when $n \ll L\tau^{-1/2}$
3. Model Specificity: Detailed analysis primarily focuses on TFIM; other models require further verification

Future Directions

1. Experimental Verification: Verify theoretical predictions on quantum simulators and quantum annealing devices
2. Interacting Systems: Generalize to strongly interacting quantum systems
3. Higher-Order Transitions: Investigate complexity behavior in first-order and topological phase transitions

In-Depth Evaluation

Strengths

1. Theoretical Rigor: Provides a complete analytical framework with rigorous mathematical derivations from Magnus operator construction to Krylov basis generation
2. Universal Discoveries: Establishes profound connections between complexity growth and known universal phenomena (KZ mechanism), possessing significant theoretical importance
3. Methodological Innovation: Non-adiabatic Magnus operator method solves technical challenges in Krylov complexity for time-dependent systems
4. Completeness of Results: Not only provides mean behavior but completely characterizes statistical properties of the distribution

Weaknesses

1. Limited Numerical Verification Range: Numerical verification is primarily confined to relatively small systems due to computational complexity constraints
2. Model Dependence: Although general arguments are proposed, detailed verification is mainly based on TFIM, a specific model
3. Experimental Feasibility: Experimental verification of theoretical predictions may face technical challenges

Impact

1. Theoretical Contribution: Makes important contributions to the intersection of quantum complexity theory and non-equilibrium quantum physics
2. Methodological Value: Provides analytical frameworks applicable to other time-dependent quantum systems
3. Practical Prospects: Possesses potential applications in quantum computing and quantum simulation

Applicable Scenarios

1. Quantum Phase Transition Research: Provides theoretical tools for studying complexity behavior in various quantum phase transitions
2. Quantum Simulation: Verify theoretical predictions in cold atom and superconducting qubit systems
3. Quantum Algorithm Design: Provides guidance for optimizing quantum algorithms based on Krylov methods

References

The paper cites 37 important references covering key works in Krylov complexity theory, Kibble-Zurek mechanism, quantum critical dynamics, and related fields, providing a solid theoretical foundation for the research.

This paper makes important contributions at the intersection of quantum complexity theory and non-equilibrium quantum physics, establishing universal connections between complexity growth and quantum phase transition dynamics, and providing new theoretical perspectives for understanding the mechanisms of complexity growth in quantum systems.