Again, none of this excuses this. It isn't an innocent mistake, which could have been caught later. The dataset is flawed and the methodology is questionable, still the authors published it on arxiv, with spectacular claims.
If you don't know, there has been a significant shift in how scientific papers (STEM for the most part) are distributed. Instead of Journals (which have lost almost all use in a digital world) papers are published freely accessible online without any formal quality control, before potentially later being published in some journal. Arxiv, where these papers are published has control over who gets to publish (not open to the public), but doesn't require a lengthy formal process. In mathematics this has worked remarkably well, notably one of the millenium problems was solved when the solution was uploaded to arxiv.
Poluting arxiv with low quality clickbait is destructive, "not being peer reviewed" is no excuse for bad science.
I think the OP is simply trying to say that a preprint cannot be "retracted".
On arxiv, it is possible to "retract" an article, in the sense that you ask that it's hidden or deleted etc. THat's not the same as retracting a published paper, where you usually get some justification, and a note from the editor explaining the decision to publish, and the decision to retract, and so on. More to the point, nobody cares if you "retract" a preprint, since it's expected that it may have errors not caught by peer-review, yet. No peer review means anyone can put anything they want online, and then take it back offline as they wish.
Note that arxiv also gives you an option to publish different versions of a paper. So you can leave your paper with errors as a v1 and upload your post-review paper, with corrections, as v2. Again, no "retraction" needed.
Different definitions of "published". Uploading a preprint to arxiv or whatever definitely counts as publishing it in the nontechnical sense of the word – to an audience comprising several billion eyeballs, no less!
Still, the question remains who published it. Some of the authors (perhaps the supervising ones) may have wished not to submit it to journals, and a zealous undergrad may have uploaded it to arxiv without removing the other authors.
