Enforcing sparseness in biologically informed shallow networks for improved interpretability Konferenzposter