Paper2Code: Automating Code Generation from Scientific Papers

(arxiv.org)

Comments

somethingsome 16 hours ago
I like the idea of having automatic code creation from papers, but I’m scared of it.

Suppose you get a paper, you automatically implement the code, and then modify it a bit with a novel idea, and publish your paper. Then somebody else does that with your paper, and does the same.. at some point, we will have a huge quantity of vibe coded code on github, and two similar papers will have very different underlying implementations, so hard to reason about and hard to change.

From a learning perspective, you try to understand the code, and it's all spaghetti, and you loose more time understanding the code than it would take to just reimplement it. You also learn a lot by not only reading the paper but reading the authors code where most of the small details reside.

And I'm not even talking about the reliability of the code, test to know that it's the correct implementation. Authors try to make papers as close as possible to the implementation but sometimes subtle steps are removed, sometimes from inadvertance, sometimes because the number of pages is lionmited.

A paper and an implementation are not one-to-one mappings

colkassad 19 hours ago
It would be neat to run their pdf through their implementation[1] and compare results.

https://github.com/going-doer/Paper2Code

wzdd 6 hours ago
I did this recently with a forward-mode AD paper, by just pasting the PDF into Claude. Like everyone, I've had mixed results with Claude coding, so I wouldn't bet my life on the output, but Claude was able to produce something for Pytorch that worked first go, had appropriate performance characteristics, and it was able to convincingly explain the connection between parts of the generated code and the paper. I was impressed.
ks2048 12 hours ago
So who has a code2paper model that we can hook up in a loop?
brundolf 9 hours ago
Not what OP is about, but idea I just had:

We should have the technology now to hand-write pseudocode on a piece of paper (or whiteboard or chalkboard), and have it translated and executed. Maybe you even hook up a projector, and project the output back onto the board

sitkack 17 hours ago
I have had good results doing bidirectional programming in Tex <=> Python.
bjourne 17 hours ago
It relies on OpenAI's o3-mini model which (I think) you have to pay for.