Physical Analogue Kolmogorov-Arnold Networks based on Reconfigurable Nonlinear-Processing Units
Kolmogorov-Arnold Networks (KANs) shift neural computation from linear layers to learnable nonlinear edge functions, but implementing these nonlinearities efficiently in hardware remains an open challenge. Here we introduce a physical analogue KAN architecture in which edge functions are realized in materia using recon...
Manuel Escudero, Mohamadreza Zolfagharinejad, Sjoerd van den Belt et al.
· Zenodo (CERN European Organi... · 0 citations