RT @SuperSTEM_UK: Here be magnons! Ground breaking study of spin wave detection at the nm-scale using HR-EELS in an electron microscope jus….

RT @BenBlaiszik: @dbngsh_mkhrj arXiv is the McMaster Carr of scientific publishing. It just works.

arXiv is more responsive and faster for both PDF (and now HTML) than any current journal website. While the former is free, and the latter charges several thousand dollars per manuscript either from the authors or from libraries.

Yet another hundredth meridian map.

RT @charliermarsh: Friday night hack: uv can now auto-detect AMD GPUs and install the appropriate ROCm PyTorch build

I have no idea what the ChatGPT model names mean anymore.

RT @hindookissinger: @puram_politics The biggest delusion that many Indians have is that their caste, their ethnic group, their sub sub sub….

RT @bscholl: This morning, I was on the road with the Symphony engine team, shopping for scalable industrial capacity to build jet engine p….

Ok. This is far from easy. Spent a long rabbit hole into how lint safe Optional types that also play nicely with JAX is elusive.

Python should have gotten a real Optional type rather than the Type | None nonsense. If I say .a: Optional[int] = 4.That should be it, rather than PyLance calling it a bug and catastrophizing about scenarios where I have passed None to a mathematical operation.

I actually like this style. from typing import Optional.aa: Optional[int] = 6. rather than the. aa: int | None = 6. I know the first is more verbose, and is the same as the second underneath the hood, but it's actually clearer in the first style that aa is an Optional variable.

RT @TheNuclearCo: U.S. Energy Secretary Chris Wright tells Congress: "America must lead the commercialization of affordable and abundant nu….

RT @Noahpinion: In the long run, what makes life good for humanity is TECHNOLOGY.

Reading public science writing is often a jarring experience. A fact is given and then an opinion follows later, which almost often has no logical step from the fact to the opinion. That opinion is at most a hypothesis if we are lucky, and yet is given with an air of complete.

So, it seems that such a project is in the works and resurrects nptyping.

One of my realizations as a scientist has been that the history of science is important too and needs to be taught. It is as essential to teach the path that science and technology took to reach here today as it is to teach other histories.

RT @Robert_Palgrave: It seems kinda sad in XPS to take the electrons out from the core orbitals. They had been there for billions of years….

RT @getjonwithit: I adore CERN. I was almost moved to tears when I first visited the LHC as a 15 year old. But if you have >$30 billion to….

ArXiv is pretty much why open science even exists. Notably multiple publishing houses vehemently opposed ArXiv. Wiley for example did not allow preprints to be posted there even a few years back.