Use plot$\_$PC in main.py to visualize the first $100$ MNIST images, as represented in the space spanned by the first $2$ principal components of the training data.

What does your PCA look like?