Structure-preserving neural operators for turbulent flows
With Imperial & industry partners
Ongoing
Research
My work spans classical physical modelling, structure-aware deep learning, and statistical mechanics—often in collaboration with academic and industry partners.
Physical & data-driven modelling
Spectral constraints in reduced-order models of geophysical fluids
2024–2026
Thermodynamically consistent learning of closure relations
Preprint in preparation
Machine & deep learning
Equivariant architectures for symmetry detection in scientific data
NeurIPS workshop submission
Stochastic processes and generative models for non-equilibrium systems
Collaboration with UCL
Statistical mechanics & stochastic processes
Work on fluctuation–dissipation relations, coarse-grained dynamics, and bridging microscopic models to macroscopic observables.