Kernel shape renormalization explains output-output correlations in finite Bayesian one-hidden-layer networks.
Sign Up to like & getrecommendations! Published in 2025 at "Physical review. E"
DOI: 10.1103/9pkk-d4bm
Abstract: Finite-width one hidden layer networks with multiple neurons in the readout layer display nontrivial output-output correlations that vanish in the lazy-training infinite-width limit. In this manuscript we leverage recent progress in the proportional limit of… read more here.
Keywords: output correlations; layer networks; hidden layer; output output ... See more keywords