KDE in Tensorflow of many points

Question

I\'m new to TensorFlow and want to make a kernel density estimate of some points. Eventually, there will be thousands of points, but I try and example with 2 here. So I make the