arXiv subjects

Ifan Williams

Publications and source records attributed to Ifan Williams.

6 recordsLinked to original sources

arXiv2026

Reinforcement Learning for Adaptive Composition of Quantum Circuit Optimisation Passes

Many quantum software development kits provide a suite of circuit optimisation passes. These passes have been highly optimised and tested in isolation. However, the order in which they are applied is left to the user, or else defined in general-purpose default pass sequences. While general-purpose sequences miss opportunities for optimisation which are particular to individual circuits, designing pass sequences bespoke to particular circuits requires exceptional knowledge about quantum circuit design and optimisation. Here we propose and demonstrate training a reinforcement learning agent to compose optimisation-pass sequences. In particular the agent's action space consists of passes for two-qubit gate count reduction used in default PyTKET pass sequences. For the circuits in our diverse test set, the (mean, median) fraction of two-qubit gates removed by the agent is $(57.7\%, \ 56.7 \%)$, compared to $(41.8 \%, \ 50.0 \%)$ for the next best default pass sequence.

quant-ph↗

arXiv2026

Deep Learning Approaches to Quantum Error Mitigation

We present a systematic investigation of deep learning methods applied to quantum error mitigation of noisy output probability distributions from measured quantum circuits. We compare different architectures, from fully connected neural networks to transformers, and we test different design/training modalities, identifying sequence-to-sequence, attention-based models as the most effective on our datasets. These models consistently produce mitigated distributions that are closer to the ideal outputs when tested on both simulated and real device data obtained from IBM superconducting quantum processing units (QPU) up to five qubits. Across several different circuit depths, our approach outperforms other baseline error mitigation techniques. We perform a series of ablation studies to examine: how different input features (circuit, device properties, noisy output statistics) affect performance; cross-dataset generalization across circuit families; and transfer learning to a different IBM QPU. We observe that generalization performance across similar devices with the same architecture works effectively, without needing to fully retrain models.

quant-ph↗

arXiv2025

A general approach to quantum integration of cross sections in high-energy physics

We present universal building blocks for the quantum integration of generic cross sections in high-energy physics. We make use of Fourier quantum Monte Carlo integration (MCI) as implemented in Quantinuum's quantum MCI engine to provide an extendable methodology for generating efficient circuits that can implement generic cross-section calculations, providing a quadratic speed-up in root mean-squared error convergence with respect to classical MCI. We focus on a concrete example of a $1 \to 3$ decay process to illustrate our work.

quant-ph↗

arXiv2024

Quantum Monte Carlo Integration for Simulation-Based Optimisation

We investigate the feasibility of integrating quantum algorithms as subroutines of simulation-based optimisation problems with relevance to and potential applications in mathematical finance. To this end, we conduct a thorough analysis of all systematic errors arising in the formulation of quantum Monte Carlo integration in order to better understand the resources required to encode various distributions such as a Gaussian, and to evaluate statistical quantities such as the Value-at-Risk (VaR) and Conditional-Value-at-Risk (CVaR) of an asset. Finally, we study the applicability of quantum Monte Carlo integration for fundamental financial use cases in terms of simulation-based optimisations, notably Mean-Conditional-Value-at-Risk (Mean-CVaR) and (risky) Mean-Variance (Mean-Var) optimisation problems. In particular, we study the Mean-Var optimisation problem in the presence of noise on a quantum device, and benchmark a quantum error mitigation method that applies to quantum amplitude estimation -- a key subroutine of quantum Monte Carlo integration -- showcasing the utility of such an approach.

quant-ph↗

arXiv2023

A Modular Engine for Quantum Monte Carlo Integration

We present the Quantum Monte Carlo Integration (QMCI) engine developed by Quantinuum. It is a quantum computational tool for evaluating multi-dimensional integrals that arise in various fields of science and engineering such as finance. This white paper presents a detailed description of the architecture of the QMCI engine, including a variety of distribution-loading methods, a novel quantum amplitude estimation method that improves the statistical robustness of QMCI calculations, and a library of statistical quantities that can be estimated. The QMCI engine is designed with modularity in mind, allowing for the continuous development of new quantum algorithms tailored in particular to financial applications. Additionally, the engine features a resource mode, which provides a precise resource quantification for the quantum circuits generated. The paper also includes extensive benchmarks that showcase the engine's performance, with a focus on the evaluation of various financial instruments.

quant-ph↗

arXiv2021

Noise-Aware Quantum Amplitude Estimation

In this article, based on some simple and reasonable assumptions, we derive a Gaussian noise model for quantum amplitude estimation. We provide results from quantum amplitude estimation run on various IBM superconducting quantum computers and on Quantinuum's H1 trapped-ion quantum computer to show that the proposed model is a good fit for real-world experimental data. We also show that the proposed Gaussian noise model can be easily composed with other noise models in order to capture effects that are not well described by Gaussian noise. We give a generalized procedure for how to embed this noise model into any quantum-phase-estimation-free quantum amplitude estimation algorithm, such that the amplitude estimation is "noise aware." We then provide experimental results from running an implementation of noise-aware quantum amplitude estimation using data from an IBM superconducting quantum computer, demonstrating that the addition of "noise awareness" serves as an effective means of quantum error mitigation.

quant-ph↗