acgan

Hi everyone. My name is Adi, and I'm a co-instructor for Stanford's CS43, Functional Programming Paradigms. We are designing our curriculum for the Winter 2020 quarter, so I would be very interested in feedback if people have any.

-AG

Heavily inspired by SICP, I am co-teaching CS43 during Stanford Winter quarter.

Code: https://github.com/stanfordhaskell/cs43 Site: https://stanford-lambda.gitlab.io

All of the course materials will be open-sourced @ https://github.com/stanfordhaskell/cs43 (previously Clojure, now Haskell!)

If anyone has concrete suggestions on how to improve our curricula, I would be happy to take any suggestions.

Thanks! We're currently discussing this with Horace.

Yup, this echoes the philosophy at e.g. Tesla. Development speed matters more than performance at first.

We really enjoyed editing this piece. Just wanted to doubly highlight a few of Horace's (chillee on HN) resources linked at the bottom:

Code: https://github.com/Chillee/pytorch-vs-tensorflow

Ablation of claims: https://thegradient.pub/p/cef6dd26-f952-4265-a2bc-f8bfb9eb1e...

JS interactive charts: https://chillee.github.io/pytorch-vs-tensorflow/

That makes sense. I'd hope that Scholar would be different, though.

A piece on how a researcher spent a summer filling out CAPTCHAs / scraping: https://www.nature.com/articles/d41586-018-04190-5

Glad to see this important piece here (disclosure: I am one of the editors of The Gradient).

https://ieeexplore.ieee.org/document/5089308, from RCIS 2009 (Beel and Gipp) noted that "Google Scholar seems to be more suitable for searching standard literature than for gems or articles by authors advancing a view different from the mainstream."

Unrelated, but interesting: scraping Google Scholar is remarkably annoying if you want to actually use the data. The easiest way (in my experience) seems to be regex hacking on the BibTeX files, but this seems truly broken.

One way to think about this is the "barbell strategy" due to Nassim Taleb. Make emails incisive and direct; decks should be contextual and comprehensive.

Thanks for sharing! Despite the fact that Shannon's "A Mathematical Theory of Communication" is so accessible, I find that most in our field (stats/ML) don't often think through information-theoretic tools in a "first principles way."

Yes, KL divergences show up everywhere, but they are not derived from scratch often enough. Maybe I'm stifled by my campus bubble though :)

Very cool! Reminds me of https://tldrlegal.com/ and https://fossa.com/.

This was really fun! Here are some thoughts:

Most researchers / students I know who are "power-users" of TeX have fairly idiosyncratic editing habits, so it would be nice to support `\newcommands` or eventually editor bindings. (I write a lot of "live-TeX" during lectures for school: https://github.com/acganesh/stanford-compendium, and I have one-character `vim` macros for each symbol). I ended up opening a `vim` pane and copying my TeX over to Chrome.

I know Evan Chen (MIT grad student) is a really prolific live-TeXer, so you might try contacting him for feedback. I'm a huge fan of this project of his: https://github.com/vEnhance/napkin.

I really like the visual equation recognition, sort of like a beefed up DeTeXify. It's very helpful to ensure that all "equivalent" TeX commands are matched appropriately.

I think there are some gotchas with space matching (I opened a few GitHub issues, e.g. https://github.com/akshayravikumar/TeXnique/issues/7), but these should be fixable.

Feel free to contact me if I can help! Sounds like a really fun project.