arXiv subjects

Tobias Farchy

Publications and source records attributed to Tobias Farchy.

3 recordsLinked to original sources

arXiv2026

Graph-theoretic design of lasing networks for physical vision

Physical neural networks perform learning through the intrinsic nonlinear dynamics of matter. Optimising their design presents a considerable challenge: complex many-body physics can provide powerful computation, but are expensive to simulate and large experimental optimisation runs are impractical to fabricate. Hence, the high-dimensional space of possible network topologies cannot be effectively directly searched. Here, we show that this search can be efficiently performed in an abstract graph space that is vastly cheaper to explore. Using random lasing networks -- composed of interconnected nanoscale waveguides and hosting strongly coupled lasing modes -- as an exemplar physical vision system, we establish a quantitative three-layer link: simple graph-theoretic metrics predict the nonlinear lasing physics, which in turn predicts vision performance. After validating this relationship using physical simulations, we exploit it to drive an evolutionary algorithm using graph metrics, producing network topologies that outperform random designs at a fraction of the computational cost (3000$\times$ speed-up compared to physical simulation). On simulated image-classification tasks, graph-optimised networks substantially improve classification accuracy. As our framework operates on network topology rather than substrate-specific physics, we anticipate it can transfer to other network-based physical learning systems, providing an efficient route for the directed design and optimisation of complex, strongly-interacting physical neural networks.

cond-mat.dis-nn↗

arXiv2025

Magnetic vortex writing and local reversal seeding in artificial spin-vortex ice via all-optical and surface-probe control

Artificial spin-vortex ice ('ASVI') is a reconfigurable nanomagnetic metamaterial consisting of magnetic nanoislands tailored to support both Ising macrospin and vortex textures. ASVI has recently shown functional applications including reconfigurable magnonics and neuromorphic computing, where the introduction of vortex textures broadens functionality beyond conventional artificial spin ice which generally supports macrospin states. However, local control of writing vortex states in ASVI remains an open challenge. Here we demonstrate techniques for field-free magnetic vortex writing in ASVI. We expand ASVI to support metastable macrospin, single-vortex and double-vortex states. All-optical writing via focused laser illumination can locally write double-vortex textures, and surface-probe writing using an MFM tip can locally write single vortex states. We leverage this writing to tailor and explore the reconfigurable energy landscape of ASVI, demonstrating programmable local seeding of avalanche-like reversal events. The global field-free texture selective writing techniques reported here expand the suite of nanomagnetic control techniques, with a host of future applications including fundamental studies of avalanche dynamics, physical memory, and direct writing of nanomagnetic 'weights' in physical neuromorphic neural networks.

cond-mat.dis-nn↗

arXiv2024

Few-Shot Neuromorphic Vision in a Nonlinear Photonic Network Laser

With the growing prevalence of AI, demand increases for hardware that mimics the brain's ability to extract structure from limited data. In the retina, ganglion cells detect features from sparse inputs via lateral inhibition, where neurons antagonistically suppress activity of neighbouring cells. Biological neurons exhibit diverse heterogeneous nonlinear responses, linked to robust learning and strong performance in low-data regimes. Here, we introduce a retinally-inspired photonic computing system where spatially-competing lasing modes in a random network laser act as heterogeneous, inhibitively-coupled neurons - enabling feature detection, few-shot classification, and segmentation. This silicon-compatible scheme harnesses heterogeneous excitatory and inhibitory nonlinear physical dynamics which give rise to emergent photonic computing behaviour, including parallel feature detection and strong performance when training data is scarce. We report 98.05% and 87.85% accuracy on MNIST and Fashion-MNIST, and 90.12% on BreaKHis cancer diagnosis - outperforming software CNNs including EfficientNetV2 and the vision transformer ViT in few-shot and class-imbalanced regimes with training sets of up to several hundred images. We demonstrate combined segmentation and classification on the HAM10k skin lesion dataset, achieving DICE and Jaccard scores of 84.49% and 74.80%. These results demonstrate the potential of random lasing networks as nonlinear photonic learning systems, and highlight the ability of heterogeneous nonlinear dynamics to support strong learning in challenging low-data scenarios.

cond-mat.dis-nn↗