Microsoft Research
3 HN recommenders · recommended in 6 months, 2018–2023
1:30:45 · 17K views · Jul 26, 2018
Conal Elliott presents his paper on automatic differentiation, deriving a simple, generalized AD algorithm from a natural specification. Varying the representation of derivatives yields forward and reverse mode algorithms that avoid the complexity and statefulness of backpropagation.
Cited as the more accessible companion to the paper, and as the most interesting treatment of AD via compiling to categories.
6 comments from Hacker News
The most interesting thing I've seen on AD is "The simple essence of automatic differentiation" (2018) [1]. See past discussion [2], and talk [3]. I think the main idea is that by compiling to categories and pairing up a function with its derivative, the pair becomes trivially composable in forward mode, and the whole structure is easily converted to reverse mode afterwards. [1]: [LINK] [2]: [LINK] [3]: Talk at Microsoft Research: this video Other presentations listed here: [LINK]
I'm yet to give reverse mode a second try, but my attempts at forward mode (linked in parent post) were on my own without the paper's help. There's a talk that Conal gives that makes the paper a bit more accessible - this video - but only a bit more. I know it'll take effort to wrap my head around the implementation details ... but the paper certainly helped clarify things I was stumbling around about.
Another paper related to AD was posted last month: The Simple Essence of Automatic Differentiation [0]. Video presentation [1]. I found it pretty awesome, just in case you hadn't seen it. [0]: https//news.ycombinator.com/item?id=18306860 [1]: this video
He also gave nice presentation on AD: this video (surprised that you didn't already mention it, considering the venue ;-))
This paper won a best paper award at ICFP 2018. Conal Elliott recently gave a longer talk about it at Microsoft Research: this video