If this were truly a blackhat effort, it could improve the quality of empirical papers at conferences specifically by practically forcing a reproducible-by-the-conference-XYZ org of the results on some benchmark using a Docker image, etc etc.<p>On the other hand, for investigative or more purely academic research, this could be another layer of nightmarish crap that people would have to sift through.<p>Also it looks like at least one example is basically cloned from the source set, just with...basically greatly reduced recall of the content (beyond some high points).<p>This feels nightmarish, to be honest. Well, I guess we'll find one way to adapt to it, or simply fall back to certain, much more provably-analog methods of doing things....<p>Real: <a href="https://arxiv.org/pdf/1409.6041.pdf" rel="nofollow">https://arxiv.org/pdf/1409.6041.pdf</a><p>Fake: <a href="http://arxivgen.com/pdfs/domain-adaptive_neur-3p5.pdf" rel="nofollow">http://arxivgen.com/pdfs/domain-adaptive_neur-3p5.pdf</a>