Near-term quantum devices are limited by noise and hardware constraints, motivating algorithmic approaches that trade circuit complexity for increased sampling overhead. Quasi-probability decompositions (QPDs), for example, allow replacing non-local operations by multiple circuits with local operations, but the associated sampling overhead generally scales exponentially and limits their practicality. In this work, we introduce a reweighting strategy for QPDs for circuits with the same variational structure across parameter settings, reusing samples and thereby reducing the sampling overhead. We first demonstrate this approach by estimating fidelities between parameterized quantum states, a key primitive in variational time evolution and quantum kernel methods. Importantly, this setup allows controlling the exponential QPD sampling overhead while preserving the structure of the state-encoding ansatz. We then apply the method to estimate the real part of the quantum geometric tensor using the simultaneous perturbation stochastic approximation and find that, in the presence of realistic hardware noise, our method outperforms other standard estimation techniques. These results highlight the potential of reweighting strategies to extend the applicability of QPD-based methods in variational quantum algorithms.
Comments
10 pages, 7 figures, 2026 IEEE International Conference on Quantum Computing and Engineering (QCE)
Near-term quantum devices are limited by noise and hardware constraints, motivating algorithmic approaches that trade circuit complexity for increased sampling overhead. Quasi-probability decompositions (QPDs), for example, allow replacing non-local operations by multiple circuits with local operations, but the associated sampling overhead generally scales exponentially and limits their practicality. In this work, we introduce a reweighting strategy for QPDs for circuits with the same variational structure across parameter settings, reusing samples and thereby reducing the sampling overhead. We first demonstrate this approach by estimating fidelities between parameterized quantum states, a key primitive in variational time evolution and quantum kernel methods. Importantly, this setup allows controlling the exponential QPD sampling overhead while preserving the structure of the state-encoding ansatz. We then apply the method to estimate the real part of the quantum geometric tensor using the simultaneous perturbation stochastic approximation and find that, in the presence of realistic hardware noise, our method outperforms other standard estimation techniques. These results highlight the potential of reweighting strategies to extend the applicability of QPD-based methods in variational quantum algorithms.
Quasi-probability decompositions (QPDs) have proven essential in many quantum algorithms and protocols -- one replaces a ``difficult'' quantum circuit with an ensemble of ``easier'' circuit variants whose weighted outcomes reproduce any target observable. This, however, inevitably yields an increased configuration variance beyond Born-rule shot noise. We develop a broad framework for accounting for and reducing this variance and prove that stratified sampling -- under ideal proportional allocation -- results in an unbiased estimator with a variance that is never worse than naïve sampling (with equality only in degenerate cases).
Furthermore, we provide a classical dynamic programme to enable stratification on arbitrary product-form QPDs. Numerical simulations of typical QPDs, such as Probabilistic Error Cancellation (PEC) and Probabilistic Angle Interpolation (PAI), demonstrate constant-factor reductions in overall variance (up to ∼60--80% in an oracle model) and robust ∼10% savings in the pessimistic single-shot regime. Our results can be applied immediately to reduce the net sampling cost of practically relevant QPDs that are commonly used in near term and early fault-tolerant algorithms without requiring additional quantum resources.
Quasiprobabilistic decompositions (QPDs) play a key role in maximizing the utility of near-term quantum hardware. For example, Probabilistic Error Cancellation (PEC) (an error mitigation technique) and circuit cutting (which enables large quantum computations to be performed on quantum hardware with a limited number of qubits) both involve QPDs. Computations based on QPDs typically incur large sampling overheads that grow exponentially with the number of error-terms mitigated or number of circuit-cuts employed, limiting their practical feasibility. In this work, we adapt the control variates variance reduction technique from the statistics literature in order to reduce the sampling overhead in QPD-based computations. We demonstrate our method using simulation experiments that mimic a realistic PEC scenario. In our experiments, we observed a more than 50% reduction in the number of samples needed to achieve a given precision, in more than 50% of the PEC-based estimations performed in the study when using our approach. We discuss how future research on constructing good control variates can lead to even stronger sampling overhead reduction.
Abstract Noisy Intermediate-Scale Quantum (NISQ) devices are near-term quantum computing devices that inherently suffer from imperfect gate implementation, decoherence, and measurement noise. The main type of variational quantum algorithms(VQAs) is based on Parameterized Quantum Circuits (PQCs), which, however, face practical challenges due to optimization instability of optimization caused by noise and reduced performance. In this paper, we suggest a comprehensive noise-resilient optimization framework for PQCs, which explicitly incorporates noise-resilient circuit architecture design, incorporating noise-aware training strategies into the variational optimization loop. The framework uses circuit structures that are efficient in depth and are hardware compatible in addition to adaptive optimizer choice, stochastic optimization, and averaging as a part of repeated measurement to reduce the impacts of realistic noise. Large-scale experiments are performed with different amounts of qubits, depths of the circuit and noise strengths with realistic noise models. The findings indicate that the convergence rate is faster, the stability is increased, the variability of run-to-run variability is minimized, and other robustness is better than the noise-unaware optimization techniques. Moreover, systematic studies determine optimal operating regions that optimally trade-off expressibility and noise tolerance to offer useful design advice on trusted implementation of VQAs with near-term quantum hardware.
Current quantum computers are plagued by prohibitive amounts of noise, which complicates the experimental realization of useful quantum algorithms that outperform classical computers.Quantum error mitigation techniques could constitute a potential avenue to demonstrate this feat without the need for fault-tolerant quantum error correction.One method in this family of mitigation techniques is the quasiprobability method introduced by Temme et al. [1].It simulates a noise-free quantum computer with a noisy one, with the caveat of only producing the correct expected values of measurement observables.The cost of a quasiprobability simulation manifests as a sampling overhead which scales exponentially in the number of error-mitigated gates in the circuit.In this thesis we aim to reduce the exponential basis of that overhead, which in turn allows the application of the quasiprobability method to deeper quantum circuits.A central result is the introduction of a novel scheme, which we call Stinespring algorithm, that aims to choose the quasiprobability decomposition in a noise-aware manner.Along the way, we introduce a generalization of the quasiprobability method, which we denote approximate quasiprobability method, that allows for a tradeoff between an approximation error and the sampling overhead.This method is already interesting on its own and we present a few potential applications.Finally, we present some ideas how quantum error correction can benefit from the quasiprobability method.This final part is separate to the other topics of the master thesis and gives a rough overview of a new research project that started towards the end of the thesis and that we will continue to pursue in the future.i Overview of ContributionsWe summarize the novel contributions from this thesis:• Many details in the presentation of Chapter 2 have not been stated in such precision in previous literature.However, we estimate that many of these contributions are evident for someone knowledgeable in the domain.• The channel difference decomposition (Section 2.6).• The approximate QPD and its applications (Chapter 3).• The Stinespring algorithm (Chapter 4).
Current and near-term quantum hardware is constrained by limited qubit counts, circuit depth, and the high cost of repeated measurements. We address these challenges for solid state Hamiltonians by introducing a logarithmic-qubit encoding that maps a system with N physical sites onto only ⌈log2N⌉ qubits while maintaining a clear correspondence with the underlying physical model. Within this reduced register, we construct a compatible variational circuit and a Gray-code-inspired measurement strategy whose number of global settings grows only logarithmically with system size. To quantify the overall hardware load, we introduce a volumetric efficiency metric that combines the number of qubit, circuit depth, and the number of measurement settings into a single measure, expressing the overall computation costs. Using this metric, we show that the total space-time-sampling volume required in a variational loop can be reduced dramatically from N2 to (logN)3 for hardware efficient ansatz, allowing an exponential reduction in time and size of the quantum hardware. These results demonstrate that large, structured solid-state Hamiltonians can be simulated on substantially smaller quantum registers with controlled sampling overhead and manageable circuit complexity, extending the reach of variational quantum algorithms on near-term devices.
We study the problem of probability distribution matching and sampling on near-term quantum computers, aiming to construct parameterized circuits that generate samples from a target distribution while minimizing resource overhead. This task arises naturally in hybrid quantum-classical workflows, where measurement-driven objectives replace full state reconstruction, and is central to applications in generative modeling and variational inference. However, it remains challenging due to hardware noise, limited circuit depth, and a high-dimensional, non-convex parameter space. We propose CircuitTree, a surrogate-guided optimization framework based on Bayesian Optimization with tree-based models for scalable, domain-aware distribution matching. Our approach introduces a structured, layerwise decomposition aligned with the variational circuit architecture, enabling distributed and sample-efficient optimization within hybrid loops with theoretical convergence guarantees. Across representative distribution-matching tasks, CircuitTree achieves up to 2-3x lower total variation distance while using 40-60% fewer gates than prior approaches. These results demonstrate its effectiveness as a practical building block for end-to-end hybrid quantum sampling.