mattkrause

6 ms·

1.
by mattkrause 4d ago
You don’t see a lot of people walking around with E-collars either, but it’s not a conspiracy. People are just way better about not biting or licking their wounds.
2.
by mattkrause 6d ago
It’s also the main reason those microwaves are so gross! Since it’s relatively thick, boiling solution sputters everywhere and is—-at best—-partially cleaned up by harried scientists. Use a big container and try not to overshoot the temp.
3.
by mattkrause 10d ago
A lot of things rely on GNSS for time synchronization, even if they're not moving....
4.
by mattkrause 12d ago
No, it's not suspicious at all! The paper was posted on the 18th of June. I would guess that means they started writing the first draft in the winter of 2026, if not earlier. 2026 data obviously wouldn't be available then, or even
5.
by mattkrause 21d ago
Nah, you can renew work permits online—-and in fact, they push you to. However, the in-country processing takes 6 months while the flagpole option (If you’re eligible) takes 15 minutes.
6.
by mattkrause 23d ago
Some of the data seems straight-up wrong. I know this computer has more RAM than reported and it's definitely not on a 4G network either.
7.
by mattkrause 1mo ago
Cool! Will check it out!
8.
by mattkrause 1mo ago
I'd also throw in hiring practice. Whatever we're doing (across the board) to find candidates seems....less ideal.
9.
by mattkrause 1mo ago
Load-bearing, even.
10.
by mattkrause 1mo ago
Statistically. You build a model that describes how likely people are to get diagnosed with Alzhimer's as a function of sex: maybe 12 women out of 100 get it, while 8/100 men are diagnosed. You can do the same thing for age: almos
11.
by mattkrause 1mo ago
It was! FWIW, my comment may have come off as more dismissive than I meant--it's a cool approach. I'd just never seen a visualization quite that like and then ran into two in the same week.
12.
by mattkrause 1mo ago
Annoyingly, the US government still sometimes does. If you want to take the patent agent exam, your CS degree has to come from an ABET-accredited program, though many of the very good ones aren't (Stanford, CMU). Likewise, federal jobs
13.
by mattkrause 1mo ago
Did someone recently release a library (or skill.md [0]) for making these sorts of visualisations? It looks strikingly like the LLM explainer someone posted a few days ago. [0] I ask this bit in the "About" makes me wonder: "
14.
by mattkrause 2mo ago
I think you'd want to specify that every element in the original list is present in the sorted list (and in the same number, in case of ties). And of course, that raises the issue of what do you want to do about ties....
15.
by mattkrause 2mo ago
The consensus was that Sandy Island was not a copyright “trap.” Some of the sightings may have been floating pumice from an undersea volcano, and the “confirmations” came from a chain of errors. https://en.wikipedia.org/wi
16.
by mattkrause 2mo ago
If the argument is that it's being manipulated, I'm not sure these stats help.
17.
by mattkrause 2mo ago
Why is that sad? It does feel like it’d be “better” if good behavior came from the company’s intrinsic behavior but laws like this help (re)establish norms that encourage those sorts of actions. Likewise, it’d be better if kids say please a
18.
by mattkrause 3mo ago
Because the customers don’t trust that you won’t just flip those switches whenever you’re planning something nefarious.
19.
by mattkrause 3mo ago
The web-routing thing is especially gross because a different example shows off pattern-matching.
20.
by mattkrause 3mo ago
It's also misses the point. People don't want "glam" papers for ego boosts and bragging rights. They want them to keep the current jobs and perhaps get better ones. Any replacement system needs to somehow serve as a toke
21.
by mattkrause 3mo ago
Some analog devices are "DSP" in a wig, but all digital signal processing is really just a convenience wrapper around analog current flow. A very handy one, admittedly, but one that might not always be worth it. There’s also a
22.
by mattkrause 3mo ago
PLOS Biology and PLOS Computational Biology are pretty well-regarded (the others are outside of my field). PLOS One does publish pretty much anything, but that was always meant to be the point: "here's some data, make of it wh
23.
by mattkrause 3mo ago
On top of that, the whole thing is done in fits and starts. You send in the final revision, it vanishes into the void for some unspecified time, and then they offer[*] you 48 hours--sometimes not even lined up with two working days!--to fig
24.
by mattkrause 3mo ago
True—-but half of them forget to actually show up and many of the ones that do try to renegotiate the deal.
25.
by mattkrause 3mo ago
And it's still true! I appreciate a birthday text/email, but a thoughtful email is nicer and an physical card/letter is better still.
26.
by mattkrause 4mo ago
It’s literally a trigram (character) language model. Check any NLP book from before 2015 or so. LLMs have more stuff bolted onto them (embeddings, RLHF) but the autoregressive core is a direct descendent of that sort of language model.
27.
by mattkrause 4mo ago
That's uhh...a language model?
28.
by mattkrause 4mo ago
Isn't this pretty much why language models were invented? Pasting something directly into the chat interface seems weird, but if you could somehow just see where P(token | context) falls off a cliff, that's a pretty good hint that
29.
by mattkrause 4mo ago
Ha! I’d never thought about it like that but…yeah. I suspect another big part of it is that marketing and sales are relatively easy to measure and to scale. You can hire one, two, or three new salespeople and expect that revenue will chan
30.
by mattkrause 4mo ago
And I’m sure those companies also have “backlogs" due to limited labor/labor costs. There are always shelves to face, vehicles with deferred maintenance, and so on. Obviously, there are limits: I’m not sure what my local grocery s

More ›