2 ms·

For papers with code, I have a seen a tendency to consider the code, not the paper to be the ground truth. If the code works, then it doesn't matter what the

by Lerc 2y ago

For papers with code, I have a seen a tendency to consider the code, not the paper to be the ground truth. If the code works, then it doesn't matter what the paper says, the information is there. If the code doesn't work, it seems like a red flag. It's not an advantage that can be applied to biology or physics, but at least computer science catches a break here.

withinboredom 2y ago
> For papers with code, I have a[sic] seen a tendency to consider the code My favorite for those is to search the code for "todo" and look up that part of the paper. Usually, these are the most complicated parts of the paper. Not always though, sometimes they are just trivial things.