# zettelkasten

## Haskell Reference

lazy, atomic/stateless, strong typing, pure (no side effects) Some pros concurrency faster What are haskell bindings? What are monads? forcing side effects into haskell What is haskell better for?

## Pyro Reference

Built with pytorch backend Usually involves specifying a generative process Usually either use Stochastic Variational Inference (SVI) optimization methods or Monte Carlo Markov Chain (MCMC) sampling methods For SVI Define a model and a guide (variational distribution) guides define where the parameters are to be learnt Example model and guide code

## Reverse a Linked List

def reverseList(self, head: ListNode) -> ListNode: current = head while (current and current.next): next = current.next current.next = next.next next.next = head head = next return head This is a real basic problem, but it can be tricky.

## Tensorflow Reference

tf.session initiates a tf graph object. Use global variable initializer with it - with tf.Session() as sess: sess.run(tf.global_variables_initializer()) ‘tf.placeholder’ Inputs to be fed in ‘tf.

## Coq Reference

Learn x in y style (* This is a comment *) (* Inductive type, enumerated finite *) (* Each item called a constructor *) Inductive day : Type := | monday | tuesday.

## Latex Reference

Omega: \omega for $\omega$ and \Omega for $\Omega$ Union: \bigcup for $\bigcup$ Infinity: \infty use \textbf{} for best practices tilde - \sim \sim

## Probability axioms

$0 \leq P(E) \leq 1$ $P(\Omega) \leq 1$ $P(\bigcup_{n=1}^{\infty}E_{n}) = \sum_{n=1}^{\infty}P(E_n)$

## Zettelkasten

German word for archive or knowledge management system

## Zettelkasten

I’m going to add another level of granularity to the notes of my website: Zettelkasten. I think there’s a strong case this will improve my learning and production. Blog posts are polished, well-researched, peer-reviewed pieces that take 15-20 hours to write.