jurgenaut23

7 ms·

61.
by jurgenaut23 2y ago
Well, it makes sense. The smaller the niche, the lesser weight in the overall training loss. At the end of the day, LLMs are (literally) classifiers that assign probabilities to tokens given some previous tokens.
62.
Interferometric single-shot parity measurement in InAs–Al hybrid devices(nature.com)
2 points by jurgenaut23 2y ago|1 comments
63.
by jurgenaut23 2y ago
Nature paper backing Microsoft's announcement of Marjorana
64.
by jurgenaut23 2y ago
As useless as it gets, surprised that it got to the front page.
65.
by jurgenaut23 2y ago
I cannot read this title without my brain bringing back memories of 30 Rock
66.
by jurgenaut23 2y ago
What kind of commentary are you referring to?
67.
by jurgenaut23 2y ago
> That neither is a solution to the actual problem is glossed over, because as a society, we've become accustomed to the idea that there is a political solution to all problems. THIS. I have started thinking a lot about this recentl
68.
by jurgenaut23 2y ago
Did you really put SQL and Spark in the same basket?
69.
by jurgenaut23 2y ago
Tokens? Is that the new « bit »?
70.
by jurgenaut23 2y ago
IBM failed pretty much at everything AI-related except for Jeopardy and Chess (and even this was brute-forcing more than learning).
71.
by jurgenaut23 2y ago
I guess they pre-trained the model to exploit common patterns found in any time-series (e.g., seasonalities, trends, etc.)... What would be interesting, though, is to see if it spots patterns that are domain-specific (e.g., the ventricular
72.
by jurgenaut23 2y ago
lol, I love this. These gems are one of the reason I hang out on HN ;-)
73.
by jurgenaut23 2y ago
Yes, exactly. It's amazing how people don't see that the _whole_ point of banks is to aggregate defaulting risks across a very large number of loans to power the economy.
74.
by jurgenaut23 2y ago
Short answer: yes, but. Without consequence, of course not. But you should be able to 'separate' your professional and personal life, even as an entrepreneur. I would NEVER found my own business if I knew that going bankrupt would
75.
by jurgenaut23 3y ago
The surprising way philosophy of science can help you become a better data scientist.
76.
by jurgenaut23 3y ago
Sure, tell yourself that.
77.
by jurgenaut23 3y ago
A quick look at the codebase seems to indicate that it doesn't. This seems very targeted at LLMs.
78.
by jurgenaut23 3y ago
Wow, it would be real fun to work on a ML-driven approach to learn the controller... and maybe do more exciting stuff than just attitude control.
79.
by jurgenaut23 3y ago
> Location based salaries are discriminatory The article is nice and well articulated, but I'd argue exactly the opposite on this point. The author seems to confuse equality and equity. He takes the Netherlands and Bay Area as an ex
80.
by jurgenaut23 3y ago
I was a super enthusiastic Quora user for a long while. This is the first time I actually took the time to _close_ an account a few months back, because I was so frustrated with the turn this whole thing took.
81.
by jurgenaut23 3y ago
I meant US/british ;-)
82.
by jurgenaut23 3y ago
Yes, bootcamps are really a great way to score high on the Dunning-Kruger score, but not so much on any other metric of value. You need lots of sweat and tears to become a capable and competent ML engineer, especially to tell apart intrinsi
83.
by jurgenaut23 3y ago
Well, even with a high level of understanding, any sufficiently advanced use case will still have some uncertainty regarding its "feasibility". Of course, you might think that some problems are "solved", e.g., OCR, trans
84.
by jurgenaut23 3y ago
I believe this is an attempt at vulgarization, as the exact opposite effect happens for anglo-saxon non-scientist readers when they come across metric units...
85.
by jurgenaut23 3y ago
We are talking about beauty here, and the first thing that comes to mind is freakin' Apple devices and Tesla vehicles? Even by HN standard, that feels like ... weird.
86.
by jurgenaut23 3y ago
Yes, and if you look at neural operators, then you're in for something that even math majors need to put a bit of effort into.
87.
HPE in Talks to Acquire Juniper(networkworld.com)
2 points by jurgenaut23 3y ago|0 comments
88.
by jurgenaut23 3y ago
Thanks for pointing this out. I will read it carefully, and maybe even order her book. Do you have a reliable review of it somewhere?
89.
by jurgenaut23 3y ago
Dunning Kruger.
90.
by jurgenaut23 3y ago
I am _not_ a doomer, by any mean, not at least in the "I'm right and you're all gonna die soon" sense. Yet, one has to acknowledge the factual correctness of this article. I also don't believe that "technology&

More ›