QoolQit
Pasqal, 2026 to present. Contributor (development and roadmap).
QoolQit is Pasqal’s open-source Python library for algorithm development in the Rydberg analog model. The idea is to let you write the algorithm once, in device-independent units, and leave compilation, unit conversion, device constraints and backend integration to the library.
Links: GitHub, PyPI, documentation
Machine learning for quantum many-body physics
CPHT, École Polytechnique and Collège de France, 2024 to 2026.
Neural-network quantum states treat a many-body wavefunction as a deep-learning model and train it with variational Monte Carlo. During my postdoc I worked on the optimisation side of that problem: natural-gradient methods and cheaper approximations of the quantum geometric tensor (the quantum analogue of the Fisher information matrix), variance reduction of stochastic gradients by importance sampling, and foundation-style models trained once across a whole family of Hamiltonians instead of one at a time. The methods are implemented in NetKet, the JAX framework for variational quantum simulation, and its extension NetKet Foundation.
Related papers: importance sampling for VMC gradients (MLST 7, 015035), block-structured approximations of the quantum geometric tensor (DiffSys workshop, EurIPS 2025), a hybrid classical-quantum algorithm for molecular spectra (arXiv:2510.24911).
Tensor networks meet the stabilizer formalism
2023 to 2025. With A. F. Mello, G. Lami, J. De Nardis, M. Collura.
Clifford circuits are cheap to simulate but not universal; tensor networks are universal but choke on entanglement. We combined the two: a Clifford layer absorbs the “easy” part of the entanglement and a matrix product state or operator handles the rest. This gave the hybrid stabilizer MPO (PRL 133, 150604) and the Clifford-dressed TDVP (PRL 134, 150403), with an application to Loschmidt echoes (PRA 111, 052401).
Benchmarking early quantum hardware
2021 to 2023. With A. Solfanelli, M. Campisi, V. Vitale, S. Gherardini, G. Giachetti.
A series of experiments on IBM superconducting devices using thermodynamic and foundational tests as benchmarks: fluctuation relations (PRX Quantum 2, 030353), Leggett-Garg inequality violations (PRA 105, 032610), qubit resetting by thermodynamic protocols (AVS Quantum Sci. 4, 026802) and measurement-induced thermalisation (J. Phys. Commun. 7, 065007).
Tensor Network Techniques for Quantum Computation (book)
SISSA Medialab, 2024. With M. Collura, G. Lami, N. Ranabhat. Open access.
A 200-page introduction to tensor networks for quantum computation and quantum information, aimed at advanced undergraduate and graduate students. Part I covers the foundations (MPS, TTN, contractions), Part II the applications.
Links: arXiv:2503.04423, publisher (DOI)