lp251

the new varia is usb-c

why are we hating on bop-it? bop-it was great fun!

why did they move to fblearner

what is the new training platform

I must know

Los Alamos is doing well. That’s about it.

This is paid for by the oil and gas boom in Southern NM.

The medical situation is getting worse by the year, though. I don’t think it’s just a matter of shoveling more dollars

shame they won't be as good

there were a bunch of reasons. couldn’t bring compiled binaries onto the red, so you had to bring the source + all deps onto a machine with no external internet.

it was unpleasant.

wonder if they still train all of their models using Mathematica because it was impossible to get pytorch on the classified systems

it does

off topic, but how is living in Estes Park?

ye, that’s a thing. don’t pluck your nose hairs.

I moved from Santa Fe County (Rancho Viejo area) to the Denver metro area and am sad every time I go outside at night. We could see the Milky Way most nights.

The NM sky is amazing during the day, too. Such a vibrant blue!

Did you ever fully acclimate? After 3 years at 6000 ft I still had lower blood oxygen than at sea level. (per apple watch, so take that with a grain of salt)

Igor, is LightOn still working on optical computing? Or have you fully pivoted to genai without optics?

"On Bandwidth"-- a paper version of a Shannon Lecture by Slepian.

The abstract begins with "It is easy to argue that real signals must be bandlimited. It is also easy to argue they cannot be so. This paper presents one possible resolution of this paradox"

http://web.eng.ucsd.edu/~massimo/ECE287C/Handouts_files/On-B...

there isn't one

a fellow longmonster!

hey, can you make some viewgraphs about the new dynamics codes?

I was born in the late 80s and C was my first language, in a community college intro to programming class.

affine, not linear. describes a line that doesn't go through the origin. that pesky shift breaks linearity

5(2z) + 2 != 2(5z + 2)

one day I’ll be able to tolerate the red at posas

not yet, though!

we’re you in kindergarten at 3? remarkable!

Mostly flash flooding in burn scars. Not really close to Albuquerque

do you remember what lounge he used for the lecture?

Ah, sorry, I can't help too much there.

I skimmed through the proofs in the SI. Many of the tools they use are taken from the compressed sensing literature (robust null space property, sqrt LASSO paths, sample complexity for CS-MRI style problems...).

I can point you to additional reading, if you'd like.

The story has changed somewhat recently, but up until ~2020 most ML for Inverse Problems research was done in Matlab, especially for MRI and CT reconstruction. Thinking of FBPNet and ADMM-Net.

This is a follow-up to an earlier paper on instabilities in NN-based solutions to inverse problems: https://www.pnas.org/doi/10.1073/pnas.1907377117

Your statements appear to be a good summary of the paper.

“integrate by parts, like you learned in middle school”

tf middle school did you go to?!

Dunno, they sure are eager to roll coal when they pass cyclists (myself included). Really cool people.

You can, but sometimes it’s not ideal.

I ran into this when A and B had many derived classes. I wanted to put A and it’s derived classes in one module, and B and it’s derived classes in another. It was messy.

I wound up putting A and B in a single module and having a separate one for the derived classes. Not ideal.