Armaan A. Abraham comments on Deep sparse autoencoders yield interpretable features too