arXiv:2610.14125
Hamiltonian-Dependent Compressibility and Physical Preservation in Approximate Quantum Compilation using Tensor Networks
Dohyun Kim, Namshik Han
Abstract
Approximate quantum compilation (AQC) using tensor networks provides a promising route to reducing the circuit resources required for quantum simulation. However, the factors governing the compressibility of many-body dynamics and the extent to which compressed circuits retain physically relevant information on quantum hardware remain poorly understood. We investigate these questions using AQCtensor for an 80-qubit one-dimensional anisotropic XYZ Heisenberg chain initialized in a N'eel state. By varying the exchange anisotropy and external magnetic field, we find that compression performance changes systematically with the anisotropic coupling configuration and field orientation. These variations are strongly associated with a nested-commutator measure related to second-order Trotter error, with corresponding directional trends also reflected in the target-state MPS bond dimension. For an 80-qubit quantum-hardware evaluation, AQCtensor compresses a 10-repetition target evolution to 5 repetitions while retaining a target-state fidelity of 0.995474, reducing the total circuit depth by 45.9% and the two-qubit gate count by 47.6%. The compressed circuit exhibits comparable behavior across magnetization, nearest-neighbor correlation, and Hamiltonian expectation measurements, while process-level diagnostics and local reduced-state tomography provide complementary characterization beyond these observables. Together, these results link AQCtensor compressibility to Hamiltonian commutator structure and target-state complexity and demonstrate substantial circuit reduction while retaining physically relevant features in an 80-qubit quantum-hardware experiment.
AI Summary
The paper examines how Hamiltonian structure affects tensor-network-based approximate quantum compilation (AQCtensor), and what physical information compressed circuits retain on hardware. It studies an 80-qubit open XYZ chain initialized in a Néel state, using MPS-based optimization of shallow Trotter ansätze against deeper target states. A parameter sweep varies anisotropic couplings and field strength and orientation at $t=0.30$, using 20-step second-order Suzuki–Trotter targets. Compression is measured by the minimum repetitions achieving $F\geq0.99$ and optimized infidelity at four repetitions. An aggregate nested-commutator measure correlates with fixed-depth infidelity (Pearson $r=0.887$, Spearman $\rho=0.900$); correlations remain positive within fixed-field-strength subsets. Its correlation with target-state MPS bond dimension is weaker ($r=0.744$). These are empirical associations, not compressibility bounds. For a separate hardware configuration on IBM Heron, a 10-repetition target is compressed to five repetitions with classically evaluated target-state fidelity $0.995474$. Transpiled depth falls from 294 to 159 and two-qubit gates from 2,487 to 1,302. Magnetization, nearest-neighbor correlations, and energy measurements compare original and compressed circuits. Four-qubit tomography gives mean hardware reduced-state fidelity $0.647$ between them. Auxiliary, shallower relative-circuit process diagnostics reach the estimation boundary at 60–80 qubits. Stated limitations include one hardware Hamiltonian and evolution time, selected local subsystems, noise-limited process characterization, restricted Hamiltonians and ansätze, and finite-budget optimization. Mitigation settings are selected separately against known reference values, making the hardware comparisons reference-guided rather than independent mitigation benchmarks. Generalization beyond these one-dimensional, short-time settings remains untested.
Error found: The stated self-mirror circuits U†U must have noiseless references <Z>=1, C_ZZ=1 and <X>=0, contradicting the substantially different Pauli-propagation references used in the mirror benchmarks. The mitigated observable estimates were selected for agreement with known reference values separately for each circuit and observable, so that agreement cannot serve as an independent validation of hardware preservation.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 20 pages, 9 figures, 5 tables; includes supplementary material