Aligned Students

Classical quantum simulations can scale and model diverse quantum environments, with which researchers can experiment new techniques and prepare for future fault-tolerant quantum hardware. This research aims to evaluate how efficiently can quantum simulations be run on classical devices, how do emerging architectures and technologies perform, and what improvements to classical architectures could benefit this research.

The project seeks to advance the understanding and capabilities of classical quantum simulation methods, as well as explore the advantages of hybrid quantum-classical hardware interaction.

My PhD research investigates representation learning in quantum neural networks (QNNs), focusing on how they dynamically align to functional modes corresponding to learned data features. I analyse the evolution of the Quantum Neural Tangent Kernel (QNTK), its eigensystem and derivatives, employing techniques from random matrix theory, perturbation theory, and quantum information geometry to describe these learning dynamics. Ultimately, I aim to develop principles for constructing QNNs that more effectively discover and utilise the underlying structure in quantum and classical data.

Neutral atom quantum computers admit promising architectures but present unique challenges in terms of their susceptibility to errors. Bespoke quantum error correction (QEC) protocols for such devices offer a path to fault-tolerant quantum computation with neutral atoms. Bringing together the design, simulation and benchmarking of advanced QEC protocols with real-time decoding on specialised hardware, my project explores the co-design of architecture, error correction codes and decoding strategies for neutral atoms. Its aim is the practical realisation of decoding on experimental setups at the UK’s National Quantum Computing Centre.

My project investigates efficient resource allocation and workload distribution across heterogeneous quantum clusters. I explore multi-model distribution strategies, enabling hybrid execution across problem, algorithm, subroutine, and model levels. Inspired by High-Performance Computing (HPC), the framework supports scalable decomposition and parallel execution of quantum programs with minimal communication overhead. A core focus is the design of mechanisms for coordinating and synchronizing distributed quantum workloads, along with early-stage exploration of distributed quantum error correction. By developing techniques for flexible, secure, and efficient orchestration of quantum resources, this research contributes to the architecture of future quantum cloud systems.

My PhD research explores the potential for quantum generative models to offer a practical advantage over classical approaches in data generation. While quantum machine learning has advanced significantly, a definitive computational edge over classical algorithms remains elusive, particularly because classical surrogates often reproduce the outputs of today’s near-term quantum devices with surprising accuracy, narrowing the expected performance gap.

Rather than focusing on label prediction, as in traditional supervised learning, my work centres on quantum generative models that aim to sample from and replicate complex probability distributions. This shift offers a fundamentally different paradigm for synthesising data. Notable early results, such as those from boson sampling and random circuit sampling, suggest that quantum devices can sample certain distributions exponentially faster than classical machines, pointing to the possibility of a genuine sampling advantage.

In my research, I investigate whether this advantage can be firmly established by exploiting nonclassical resources, such as contextuality, entanglement, and other resource-theoretic constructs, as the foundation for sampling tasks that classical systems struggle to replicate efficiently. By grounding quantum generative modelling in these uniquely quantum phenomena, my goal is to move beyond theoretical promise and toward concrete, practical benefits in data generation.

As an Industrial Doctoral Landscape Awards (IDLA) student with the QSL and NQCC, I will be developing rigorous protocols to verify and benchmark quantum computations.

As part of the NQCC’s Testbed programme, I will work with industry-developed experimental quantum platforms to fairly evaluate their performances as candidates for future fault-tolerant quantum computing. With the QSL at the University of Edinburgh, I will also further existing verification protocols to ensure the cryptographic security of future fault-tolerant quantum computations performed between multiple parties.

Previously, I researched rare particle decays at LHCb as part of my MMathPhys degree at University of Warwick.

My research focuses on quantum low-density parity check (qLDPC) codes, specifically, their compilation to the circuit level, the development and improvement of their decoders, and their mapping and optimisation for ion trap devices.

My PhD research focuses on developing fault-tolerant architectures for near-term quantum computers, with a particular emphasis on applying quantum low-density parity-check (qLDPC) codes to neutral atom platforms. These high-rate codes promise scalable error correction with reduced overhead, making them attractive for emerging (dual-species) Rydberg tweezer arrays. I investigate how such architectures can support real-time quantum error correction by characterising hardware-specific noise and connectivity. Alongside this, I co-develop software tools for simulating and benchmarking QEC performance under realistic conditions, aiming to bridge the gap between theoretical code design and practical implementation on quantum hardware.

My research focuses on the co-design of quantum programming languages and their compilers.

The architecture for quantum computers admits an inherent notion of parallelism and distribution that should be handled natively in quantum programming languages themselves. The constructs that we provide for this should both allow modular composition of concurrent programs which can be used to optimise their compilation to an actual quantum computer.

If concurrent programs and distribution of these should be core primitives of quantum programs, what do models of these look like, how can we give them a semantics, and how do they relate their classical analogues?