Still trying to figure out the best way to use the affordances of this platform for discussing papers on #arXiv (that aren't my own). I went with a content warning to create a subject line of sorts, and screencap of the first page.

But maybe next time I won't use the CW function.

@Riedl CW as content “hint” (to make the feed more scannable) is a nice one. But should be related to content not category yeah? Ie, not “Arxiv Paper” but “Arxiv Paper about (…)” ? I feel like these affordances are going to get hashed out eventually! :)
@CyberneticForests Warning: this arxiv paper might shatter your beliefs about how transformers work.
@Riedl Helpful CW for any convolutional neural networks who might be reading