Thanks! It's great to have both links, but in cases like this where the material is highly specialized (and not on a topic like computing that the audience here is familiar with), it's usually better to submit the best available popular article and then link to the paper in the comments. I've switched to that from https://arxiv.org/abs/2210.14154 now.
the title implied something potentially really interesting to me that is not actually in the article: I thought "from Distant Reactors" meant detecting extraplanetary intelligience capable of building reactors; I'm not sure what it does mean, but I guess it's earthbound reactors at some distance from the experiment.
The near / distant difference is something that's known to physicists but could be misleading to the public, as you have experienced.
There is a long-standing question about the neutrino spectrum that comes out of reactors, the "reactor neutrino anomaly", which is a ~6% mismatch between a calculation of how many electron (anti)neutrinos should come out of reactors and how many are observed. Since neutrinos change their flavor, it's a potential avenue for observing so-called sterile neutrinos or discovering a fourth generation of leptons.
The neutrinos observed far from the reactor have a longer flight and therefore more time to change flavor; you'd expect a good mixture. That's what makes a detector "distant" (or "long baseline").
The neutrinos observed "close" to the reactor, in contrast, have less time to mix and you expect to see primarily electron flavor. To maximize discovery potential for a fixed detector sensitivity, close winds up meaning in practice as close to the reactor core as logistics and safety allow (~10 meters).
