5 papers
Reservoir Computing with Heterogeneous Magnetic Metamaterials
R. Yagan, C. Swindells, I. T. Vidamour +6
Physical reservoir computing utilizes the intrinsic nonlinear and history-dependent dynamics of physical systems to perform machine-learning tasks with minimal training overhead. H…
Low-power analogue neural networks with trainable nonlinear connections for continuous control
Ian T. Vidamour, Fernando Aguirre, Thomas J. Hayward +13
Physical neural networks promise low-power machine learning by computing directly with analogue device physics, but most architectures force nonlinear device responses to act as sc…
Learning Nonlinear Heterogeneity in Physical Kolmogorov-Arnold Networks
Fabiana Taglietti, Andrea Pulici, Maxwell Roxburgh +10
Physical neural networks typically train linear synaptic weights while treating device nonlinearities as fixed. We show the opposite - by training the synaptic nonlinearity itself,…
RingSim- An Agent-based Approach for Modelling Mesoscopic Magnetic Nanowire Networks
Ian T Vidamour, Guru Venkat, Charles Swindells +9
We describe 'RingSim', a phenomenological agent-based model that allows numerical simulation of magnetic nanowire networks with areas of hundreds of micrometers squared for duratio…
Noise-Aware Training of Neuromorphic Dynamic Device Networks
Luca Manneschi, Ian T. Vidamour, Kilian D. Stenning +13
Physical computing has the potential to enable widespread embodied intelligence by leveraging the intrinsic dynamics of complex systems for efficient sensing, processing, and inter…