Recent Publications

On the Computational Complexity of Blind Quantum Machine Learning · arXiv 2026
Can you get a quantum speedup and keep your data private? We know the answer is yes! Using MasonQ, our universal brickwork transpiler (and also company: learn more at MasonQ.io), we examine blind versions of the QFT, HHL, quantum recommendation systems and even quantum transformers for blind quantum artificial intelligence.

High-Fidelity Remote Graph State Preparation for Blind Quantum Computation · arXiv 2026
Why send qubits one at a time when a single photon can carry a whole graph state? Our protocol encodes qubits and their entanglement structure into one high-dimensional photon, so the server barely has to entangle anything. Fidelity doesn’t depend on graph shape, and a six-qubit QFT shrinks from 420 remotely prepared qubits to 48.

Partially-Blind Single-Qubit Classification over a Prototype Hybrid Quantum Network · arXiv 2026
This is blind quantum machine learning in its smallest form: a classifier that runs privately on a remote server with just one qubit on each side. It’s built for the quantum networks we have today. It’s a first step toward experiments where quantum learning and quantum privacy meet.

Meeting the Needs of the Global Quantum Science Community: A Call to Action · arXiv 2026
Quantum mechanics turned 100 in 2025, and its community still doesn’t look like the world. With DiviQ and the Quantum Flagship EDI Working Group, we ran the largest international surveys yet on diversity, climate and opportunity in quantum. More than 1,000 people in 36 countries responded, and half of marginalized respondents have considered leaving the field. The data are open, and the need for action is urgent.

Towards the Geometry of Quantum Optical Measurements: Kraus Trajectories in the Instrumental Weyl-Heisenberg Group Manifold · arXiv 2026
In the spirit of Lie and Klein, who put transformation groups at the foundation of geometry, Chris Jackson and I study continuous measurement in measurement’s own terms. Kraus operators trace trajectories through the instrumental Weyl–Heisenberg group, independent of any quantum state. Introducing Iwasawa coordinates to quantum optics, we show that the canonical projection reveals a fiber bundle: a base manifold of positive operators with post-measurement unitaries as fibers. This sets up parallel transport of POVM elements and a genuine geometry of quantum measurement, enables us to describe (for the first time) adaptive measurement in a state-independent formalism, and lays the foundation for a new generation of tailor-made measurements.