Abstract Quantum noise limits the reliability of noisy intermediate-scale quantum (NISQ) devices. Quantum error mitigation (QEM) provides a near-term path to improved accuracy, with randomized compiling (RC) serving as a key technique for transforming arbitrary gate noise into a stochastic Pauli form. However, its sampling complexity grows exponentially with circuit depth, hindering practical use. We introduce twirling scheduling (TS), which groups gate cycles into configurable units twirled collectively. By incorporating pseudo-twirling for non-Clifford gates, TS reduces the sampling-space dimension, mitigating this exponential scaling. Our theoretical analysis provides rigorous error bounds. Numerical simulations on Quantum Approximate Optimization Algorithm circuits up to 50 layers show that TS reduces estimation error by approximately 62% compared to conventional RC under a very small sampling budget and maintains superior performance when combined with established QEM protocols, including zero-noise extrapolation, Clifford data regression, and virtual distillation. TS thus offers a scalable, practical advance for error-mitigated quantum computation on NISQ hardware.
Abstract Quantum noise limits the reliability of noisy intermediate-scale quantum (NISQ) devices. Quantum error mitigation (QEM) provides a near-term path to improved accuracy, with randomized compiling (RC) serving as a key technique for transforming arbitrary gate noise into a stochastic Pauli form. However, its sampling complexity grows exponentially with circuit depth, hindering practical use. We introduce twirling scheduling (TS), which groups gate cycles into configurable units twirled collectively. By incorporating pseudo-twirling for non-Clifford gates, TS reduces the sampling-space dimension, mitigating this exponential scaling. Our theoretical analysis provides rigorous error bounds. Numerical simulations on Quantum Approximate Optimization Algorithm circuits up to 50 layers show that TS reduces estimation error by approximately 62% compared to conventional RC under a very small sampling budget and maintains superior performance when combined with established QEM protocols, including zero-noise extrapolation, Clifford data regression, and virtual distillation. TS thus offers a scalable, practical advance for error-mitigated quantum computation on NISQ hardware.
The conventional circuit paradigm, utilizing a small set of gates to construct arbitrary quantum circuits, is hindered by significant noise. In the quantum Fourier transform, for instance, the standard gate paradigm employs two CNOT gates for the partial CPhase. In contrast, some quantum computers can directly implement such operations using their native interaction, resulting in less noisy gates. Unfortunately, coherent errors degrade the performance of these gates. In Clifford gates such as the CNOT, these errors can be addressed through randomized compiling (RC). However, RC does not apply to the non-Clifford multi-qubit native implementations described above. The present work introduces and experimentally demonstrates a technique called ‘Pseudo Twirling’ (PST) to address coherent errors. We demonstrate experimentally that integrating PST with the ‘Adaptive KIK’ quantum error mitigation method enables the simultaneous mitigation of noise and coherent errors in multi-qubit non-Clifford gates.
Quantum error mitigation (QEM) schemes have greatly enhanced the performance of quantum computers, mostly by reducing errors caused by interactions with the environment. Nevertheless, the presence of coherence errors, typically arising from miscalibration and interqubit crosstalk, is a significant challenge to the scalability of quantum computing. Such errors are often addressed using a refined Pauli twirling scheme called randomized compiling (RC) that converts the coherent errors into incoherent errors that can then be mitigated by conventional QEM. Unfortunately for multiqubit gates, RC is restricted to Clifford gates such as cnot and cphase. However, it has been demonstrated experimentally that a direct implementation of multiqubit non-Clifford gates, i.e., without using multiqubit Clifford gates, has reduced the depth of the circuit by a factor of four and more. Recently, a framework called pseudotwirling (PST) for treating coherent error in multiqubit non-Clifford gates has been introduced and experimentally demonstrated. We show analytically that a higher-order correction to the existing PST theory yields an over-rotation coherent error generated by the PST protocol itself. This PST effect has no analog in RC. Although the small induced over-rotation can amount to a significant coherent error in deep circuits, we explain why it does not degrade the performance of the gate. Interestingly, we find that a simplified twirling scheme that was introduced and exploited experimentally by Kim et al. also displays an induced over-rotation. We study the conditions under which the two twirling schemes display the same over-rotation behavior.
Simulating real-time dynamics under a Hamiltonian is a central goal of quantum information science. While numerous Hamiltonian-simulation quantum algorithms have been proposed, the effects of physical noise have rarely been incorporated into their performance analysis, despite the non-negligible noise levels of quantum devices. We present an end-to-end complexity analysis of noisy Hamiltonian simulation combined with quantum error mitigation (QEM) to answer how many circuit runs are required to reach a given target accuracy. Because the QEM sampling overhead grows exponentially with the circuit depth while the algorithmic error decreases with it, the circuit depth becomes an optimization variable, and we derive an analytic depth-selection rule for two algorithm families. For the order-k Suzuki--Trotter formula, the optimized cost exhibits a critical error εc, below which the required number of circuit runs grows exponentially. QEM improves the noise dependence of εc from sublinear to kth-power scaling, an exponent improvement by a factor of k+1. For randomized-LCU-based simulation, optimizing the repetition number yields a square-root improvement in the simulation-time dependence of the sampling-overhead exponent over the standard parameter choice. We further quantify the noise-characterization cost required for error mitigation via gate set tomography and the recently proposed space-time noise inversion method, showing that the latter can significantly reduce this cost.
Group twirling is crucial in quantum information processing, particularly in randomized benchmarking and randomized compiling. While protocols based on Pauli twirling have been effectively crafted to transform arbitrary noise channels into Pauli channels for Clifford gates—thereby facilitating efficient benchmarking and mitigating worst-case errors—the lack of practical twirling groups for multiqubit non-Clifford gates remains a challenge. To address this gap, we study the issue of finding twirling groups for generic quantum gates, focusing on a widely used circuit structure in randomized benchmarking or randomized compiling. Specifically, for multiqubit controlled phase gates, which are essential in quantum algorithms and directly implementable in practice, we determine optimal twirling groups within the realm of classically replaceable unitary operations. Contrasting with the local Pauli twirling group for Clifford gates, the optimal groups for such gates contain nonlocal operations, highlighting the overhead of tailoring noise in global non-Clifford contexts. We design new benchmarking procedures for multiqubit controlled phase gates based on the optimal twirling groups. Our simulation results show that our scheme can improve the precision and accuracy of benchmarking in small-scale systems. Published by the American Physical Society 2024
Measurement error mitigation (MEM) is essential for realizing reliable quantum computation. Model-free measurement error mitigation (MF-MEM) is an important class of MEM methods that employs Pauli twirling—typically with a random twirling set—to convert measurement noise into a state-independent scaling factor, thereby enabling error mitigation through simple calibration. However, such methods face prohibitive sampling overhead, limiting their scalability. To address this, we introduce subsystem-balanced Pauli twirling (SB-PT), a twirling method designed for MF-MEM that enforces Pauli operator balance on measuring subsystems to selectively suppress dominant measurement noise. Theoretically, for a weight- τ Pauli observable, SB-PT removes all independent error components using only O ( 4 τ ) random circuits, substantially reducing the sampling overhead over conventional Pauli twirling. This efficiency gain is most significant for sparse observables. To extend such resource-efficient mitigation to arbitrary observables, we generalize and integrate a hardware-efficient measurement transformation framework that converts high-weight Pauli operators into low-weight effective ones via linear-depth circuits. The circuit noise introduced during this transformation is jointly mitigated with native measurement noise using a unified twirling protocol, ensuring robust performance. Extensive numerical simulations demonstrate a greater than 16-fold improvement in sampling efficiency over conventional random twirling, with consistent performance gains across varying system sizes and error regimes. This work provides a resource-frugal and experimentally viable path toward high-fidelity measurement in near-term quantum devices.
The Quantum Approximate Optimization Algorithm (QAOA) is widely regarded as a promising near-term quantum algorithm for addressing hard combinatorial optimization problems on Noisy Intermediate-Scale Quantum (NISQ) devices; however, its real-world effectiveness is severely constrained by hardware noise, including gate infidelities, decoherence, and measurement errors. This work explores an error mitigation–driven approach to enhance the performance and robustness of QAOA without altering its algorithmic structure or requiring fault-tolerant quantum resources. Measurement error mitigation and zero-noise extrapolation are incorporated into the execution and post-processing stages of QAOA circuits and evaluated using the Minimum Vertex Cover problem as a representative benchmark. Performance is assessed through approximation ratios, solution stability, and noise sensitivity across multiple circuit depths. The results indicate that error mitigation not only improves approximation accuracy but also stabilizes output distributions and delays performance degradation at increased circuit depths. Additionally, the mitigated executions demonstrate better consistency across repeated runs, highlighting improved resilience to stochastic noise. These findings emphasize that lightweight and hardware-efficient error mitigation techniques can significantly extend the practical usability of QAOA on current NISQ hardware, providing a scalable and realistic pathway for executing quantum optimization algorithms more reliably in near-term quantum computing environments.