Occasionally, when major papers get written, they somehow end up leaked, such as when Google's (now published) Sycamore paper got temporarily posted on NASA's website.

While general questions may be OK, specific implementation questions, or questions that focus on such papers may be premature.

Should we take this approach? Or are our policies broad enough to address such questions anyway?


With two downvotes it seems the consensus is to not be inhibited from asking/answering!!

  • $\begingroup$ What would be the advantage of not asking questions based on the Syracuse paper? Or do you mean asking questions on the temporarily posted copy instead of the fully copy which will presumably be published soon? $\endgroup$ – heather Sep 23 '19 at 13:58
  • $\begingroup$ The latter... The advantage of not asking questions is that it might help avoid a pile-on of questions that no one is prepared to answer now. $\endgroup$ – Mark S Sep 23 '19 at 14:45
  • $\begingroup$ They did publish a paper a while back about demonstrating quantum supremacy, so questions about that are definitely OK. Everything else, though... $\endgroup$ – Mithrandir24601 Sep 23 '19 at 15:12
  • $\begingroup$ It’s probably gonna drive traffic to ask questions like what are the gates used in Sycamore. Also Aaronson has been free to answer high level questions. I’m not familiar enough with the reddit thread. Maybe our policies are broad enough to handle most technical questions, but questions like why was it prematurely released are not Germaine? The biggest question that I have is why 53 when IBM and Alibaba have said 53 is classically simulable? $\endgroup$ – Mark S Sep 28 '19 at 20:24
  • 1
    $\begingroup$ Well, it's also for the sake of both politeness and the fact that the people who know most about it (i.e. anyone we have here who's done actual work on this, of which there's definitely at least one) likely won't be able to answer any questions until it's published (or had a press conference etc.) $\endgroup$ – Mithrandir24601 Sep 29 '19 at 15:44
  • 1
    $\begingroup$ The paper has now been published! - as this question as it stands is now outdated I was going to update it to be about leaked papers in general, so that it stays relevant, if you're OK with that? $\endgroup$ – Mithrandir24601 Oct 23 '19 at 9:32
  • $\begingroup$ @Mithrandir24601 yes! $\endgroup$ – Mark S Oct 23 '19 at 11:18

I saw this just now, and I think it is more or less established that we are fine with having these questions on the site by now. I'll add my opinion as an answer just to not leave this question hanging.

I see no problem with discussions about the Sycamore paper, as most of the details will probably be unchanged when the paper actually comes out, and the current version is effectively available online for reference. And even if some detail that was the subject of some question here does end up being different in the published paper (which will probably not happen anyway), this could be an interesting point of discussion of its own.

  • 1
    $\begingroup$ I've made a relatively significant edit to the question to make it more general, given the Sycamore paper is now published. I think this answer still stands on its own without requiring any editing but just letting you know in case you want to adjust anything $\endgroup$ – Mithrandir24601 Oct 23 '19 at 17:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .