Self-classifying MNIST Digits

1. 1Define the cellular automata architecture
2. 2Implement end-to-end differentiability for training
3. 3Train the model on MNIST dataset
4. 4Evaluate the model's performance on test data

ML researchers and engineers working on innovative classification models can benefit from this approach, as it offers a unique perspective on traditional classification tasks

💡 Cellular automata can be used for image classification tasks with end-to-end differentiability

🤖 Self-organising cellular automata for MNIST digit classification