rsfern

15 ms·

181.
by rsfern 3y ago
> Just calling it an array might be underselling it, at this point I like the NDArray terminology used by numpy. I think it gets the point across more clearly than “tensor”, and conceptually they’re pretty much equivalent
182.
by rsfern 3y ago
Sometimes that is included in the journal LaTeX template, but otherwise authors (in my field at least) typically won’t say where a paper is submitted before it’s accepted for publication there, or at the very least if it’s made it past the
183.
by rsfern 3y ago
This is quite the hot topic in chemistry and materials for the last few years. See https://arxiv.org/abs/2304.11120 for a current perspective-ish review
184.
by rsfern 3y ago
Your ipython profile suggestion is good, I use that for `%autoreload` so I don’t see why it wouldn’t work for ipyflow interesting question, I’m going to have to try the opt in reactivity in ipyflow because it’s not an option in Pluto. Actua
185.
by rsfern 3y ago
Pluto is awesome Looks like by default you have to manually trigger reactivity in ipyflow, but there is a `%flow mode reactive` ipython magic mode that enables Pluto-style reactivity!
186.
by rsfern 3y ago
Do you have any prompt tips for this sort of task? I’ve tried it a bit ago and didn’t manage to get very good results even with a lot of iteration. Could have been the niche subject matter though
187.
by rsfern 3y ago
Figures 9 and 12 for anyone else who wants to scroll straight to the trippy diffraction patterns they computed
188.
by rsfern 3y ago
The way you present the distinction is really clear, but in the physical sciences it’s very common to use the terms AI and ML interchangeably, and I think it’s partly because people in those fields are not quite sure of the distinction betw
189.
by rsfern 3y ago
Ironically, attention is what you need to get high citation counts Whether quirky titles help or hurt with that i think depends on name recognition and the publication venue
190.
by rsfern 3y ago
The scale factor is 9/5, which is close enough to 2 that the precision is fine for casual conversion (weather report, discussion of boiling tea on the web, …)
191.
by rsfern 3y ago
USian here, two perspectives. I wish we used the SI system, it’s clearly superior and I agree about the communication problems. It’s also really not that hard to convert temperature roughly, just a factor of two and a constant.
192.
by rsfern 3y ago
Are you a researcher or a citizen scientist? Either way it might be worth reaching out to one of the beamline scientists who runs a measurement technique you might be interested in. Usually you would chat with them to discuss the measuremen
193.
by rsfern 3y ago
NIST doesn’t make any regulations, it’s a standards and metrology agency. Is there something specific in this framework that you think benefits big tech (or some other parties) but is not in the best interest of the public?
194.
by rsfern 4y ago
NIST isn’t a regulatory agency, their whole mission is basically precompetitive consensus building like sibling comment mentions Also, there’s a nontrivial amount of AI research done by NIST scientists
195.
by rsfern 4y ago
I’m not sure about resuming from sleep, but does mosh address your network stability issues? https://mosh.org/
196.
by rsfern 4y ago
I’m not deeply familiar with all these papers, but two things stand out to me The model architectures are different, and in the very latest paper they scale these not-transformer models to sequence length of 64k, where the paper you linked
197.
by rsfern 4y ago
Yes, I agree with the sibling poster that it’s a pretty straightforward transition, especially without array broadcasting baggage. Personally I think programming in Julia is usually more fun than python, and I think my python code has also
198.
by rsfern 4y ago
I love org mode for scientific writing (especially with org-ref [0]) but it’s just not very convenient for collaborative projects because there are very few emacs users (in my field at least). Overleaf has been workable in my experience, bu
199.
by rsfern 4y ago
Generally agree, but if you write pythonic Julia it can lead to performance issues (in particular for numerical code, which I realize is not the focus of this post). It’s taken me a while to unlearn the numpy/torch style of heavily arr
200.
by rsfern 4y ago
In Julia you can use https://juliaarrays.github.io/StructArrays.jl/stable/ which lets your code use the array of structs interface but internally it’s a structure of arrays
201.
by rsfern 4y ago
DOE already funds a ton of research programs in battery tech, PV, and other renewables. I mean it would be great if they could fund even more, but you make it seem like some kind of politically untenable topic when it’s not
202.
by rsfern 4y ago
That’s cool. I was wondering how this compares to ray (which I use with my institutions slurm-based clusters). The scheduler system that determined.ai has seems a lot more granular which suits the workloads you get with a team of people doi
203.
by rsfern 4y ago
Not quite chemistry, but “Stuff Matters” by Mark Miodownik is a really good accessible introduction to a lot of foundational materials science ideas
204.
by rsfern 4y ago
I think you’re right, it’s unused. It’s not listed in the nodes below, so I’m not sure if it’s left in to illustrate that the tracer can tell it’s unused. It’s not called out though, and it has the same shape as the first example, so maybe
205.
by rsfern 4y ago
Thanks for sharing your experience, I think that adds a lot more to the discussion than your previous comment, which frankly came off kind of callous and out of touch because it reads like you’re writing off “kids these days”. I don’t much
206.
by rsfern 4y ago
Thanks, had dark mode on so didn’t notice the green username, but that does make more sense of the tone of the comment
207.
by rsfern 4y ago
Today I used chatGPT to workshop title ideas for a research paper I’m finalizing. (It wasn’t able to produce good backronym titles, sadly, but it was a pretty good source of feedback/ideas) Some graduate students I collaborate use it
208.
by rsfern 4y ago
Do you know any of these people personally, or is that just what it seems like (based on what you’ve seen on the web or something?). I know a few trans and non-binary people and they don’t really fit the picture you paint (admittedly also a
209.
by rsfern 4y ago
That’s a cool team setup! As a fellow computational scientist without much formal CS/SwEng background, do you mind sharing any insights you’ve found valuable so far?
210.
by rsfern 4y ago
Lots of materials science interest in this area. Applications are pretty diverse, but there are definitely opportunities at the microscope OEMs, as well as in manufacturing - steel and metals processing like sibling mentioned, also aerospac

More ›