Repository owner: Ned Taylor (nedtaylor)
Neural networks with convolutional, graph-based and physics-informed layers.
Repository · Pinned source (4b81aeac0711) · Try in playground · Checked package example
Topics: automatic-differentiation, cnn, convolution, convolutional-neural-networks, deep-learning, feed-forward-neural-networks, fortran, fortran-package-manager, graph-neural-network, inverse-design, machine-learning, message-passing-neural-network, neural-network, neural-networks, physics-informed-neural-networks
Neural-network function approximation. A fixed-weight dense layer with tanh activation is compared against direct Fortran matrix multiplication and TANH. This does not test training, automatic derivatives, graph convolutions or model accuracy on a dataset. K remains broad.
Resolved GitHub repositories used by the verified playground build, combining direct and indirect dependencies. This list does not describe internal module dependencies or native system libraries.
Build record verified 2026-10-28.
No other selected build records this repository as a dependency. This is not a measure of usage outside the catalog.
These mappings have pinned source evidence. A checked package example may exercise only some of them; source review is not a numerical benchmark.
Module: athena__full_layer
K: verified
Derivatives / callbacks: not applicable
Neural-network function approximation. A fixed-weight dense layer with tanh activation is compared against direct Fortran matrix multiplication and TANH. This does not test training, automatic derivatives, graph convolutions or model accuracy on a dataset. K remains broad.
Pinned source: src/athena/athena_full_layer.f90
Author examples/tests: test/test_full_layer.f90