Hopfield Networks in C

A simple implementation of traditional Hopfield Networks (w/ binary states) in C. A two-class binarized subset of MNIST is used as a toy example.

How to Use This Code?

1. Run the generateData.py script to load and preprocess the example data.
2. Run make or make debug to compile the code (Note: Compiling the code in DEBUG mode will print the network state and energy at every training iteration.).
3. Run the binaries in ./HopfieldNetworks

Ressources

• "Neural networks and physical systems with emergent collective computational abilities" (Seminal Paper)

• Wikipedia Article on Hopfield Networks

• "Hopfield Networks is All You Need" (Paper)

• Good Blog Reference on Hopfield Networks