Making neural networks more interpretable, for research and art :mag_right: :computer: :brain: :art: - View it on GitHub