This sentence -- not to mention the rare pepe collection at the end -- made me pause a bit and wonder if the paper isn't a particularly elaborate chan troll. Clearly "rare pepe" is a chan in-joke.
Rare Pepe is the "name" of the meme as far as the subject specific custom pepe images/shops go.
But I still can't believe that there is a research paper out there where there is a section which describes the researcher's "collection" of rare pepe memes with visual aide samples.
I can't believe that a presidential candidate prominently featured a page on their website attacking a stupid internet meme. I think the Aztecs fudged their calculations somehow, and it was really supposed to be 2016...
Remember that arXiv is just pre-print and the last page if not the entire paper is probably nothing more than a lame/distasteful joke. Remember that any nutbag who knows a little English and LaTeX can publish on arXiv. It will likely not make it into a journal.
Seeing "LaTeX" (lay-tech) and "arXiv" together in your sentence made me finally realize that arXiv is pronounced "archive". I've been reading it as "arx-iv" for years.
Knuth, in his book, memorably decreed a pronunciation (and case mixing in the written form) for TeX, with a hard /ch/ at the end (literally, "your monitor should become slightly moist"), backed by a clever and erudite rationale involving the etymology of words like technique.
Leslie Lamport, in his book on LaTeX, decreed a case mixing, but (as you say) ok'd any reasonable pronunciation.
It is not entirely a free-for all (as in, anyone can submit papers), there is an endorsement system [1]. Of course, it's only networking and who-knows-who - but there is controversy, as in established contributors think that the endorsement system is too high a barrier.
