Neural Networks, Types, and Functional Programming

I really liked the mention of the "three narratives" of neural networks: 1) human brain 2) nonlinear "squashing and folding" of the input space 3) probabilistic / latent variables. Seeing different ways to look at something is always refreshing and opens up deep learning to rich analytical techniques developed in optimization, type theory, etc. HN: how do YOU think about NNs? As a matter of pure preference, most of my daydreaming of NNs comes from my inductive biases of (1)…

Folks who are into this may also like the "Type Safe Neural Networks in Haskell" https://blog.jle.im/entry/practical-dependent-types-in-haske... I'm also starting work on a set of bindings to libdarknet for Idris with similar properties.

When one interprets types as spaces, as in homotopy type theory, the correspondence between types, representations, and the manifold hypothesis not only become real, they become programmatic and explocitly representable. This incredible superpower is something humankind is only beginning to grasp.

I wanted my first comment on HN to be on an Article where I really learned something. Your article on Neural Networks, Types, and Functional Programming did just that. Thanks for taking the time to write it.