Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Imnimo
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
77 ms
·
961.
▲
by
Imnimo
5y ago
My prime number skills only go up to 47 because that's how high Number Munchers went.
962.
▲
by
Imnimo
5y ago
So, if I understand correctly, this is not an actual measurement of urban heat, but rather is the prediction of a model which takes as input different types of land cover in the cities? It sort of seems like they should just measure the tem
963.
▲
by
Imnimo
5y ago
My personal heuristic on reward curves is the opposite - if they're separated it probably means the baseline is weak rather than that the new method is revolutionary. Close curves probably means the paper is comparing to a strong sota
964.
▲
by
Imnimo
5y ago
I'm not making any claim about what the null hypothesis should be here. I'm simply disagreeing that the datapoint says what the poster claims it says.
965.
▲
by
Imnimo
5y ago
All I'm saying is that your statistic does not disprove the allegation that Google's hiring processes might favor white workers. You're making an apples-to-oranges comparison, and drawing a conclusion that isn't demonst
966.
▲
by
Imnimo
5y ago
>with the assumption that they'd fail and eventually have to be let go. This seems at odds with the article: >Nearly all residents converted to regular employees, according to the presentation
967.
▲
by
Imnimo
5y ago
Google hires many people who are not from the US to come work in the US, so I don't think the 61.5% is reflective of their hiring pool.
968.
▲
by
Imnimo
5y ago
From a Bayesian perspective, I think we should consider the luck required for this hypothesis as a strike against it. In other words, if the hypothesis contends that observing an Einstein was very unlikely, then the fact that we did observ
969.
▲
by
Imnimo
5y ago
It seems like we'd have to have been extremely lucky for the one Einstein to have been born where he was and not as a subsistence farmer if there were zero potential Einsteins among the subsistence farmer population.
970.
▲
by
Imnimo
5y ago
I wonder if you could make the noise inputs work again by using the same process as for the latent code - generate the noise in the frequency domain, and apply the same shift and careful downsampling. If you apply the same shift to the noi
971.
▲
by
Imnimo
5y ago
Here is a simple implementation I made a while back with plain-english comments rather than formulas, if that's helpful. https://github.com/nemtiax/CFR/blob/main/cfr.py
972.
▲
by
Imnimo
5y ago
I'm curious what the probability of contracting the allergy is from a single tick bite? Is it the sort of thing where the chances are low, but if you live in an area or engage in activities where you're getting a lot of bites, yo
973.
▲
by
Imnimo
5y ago
>I understand why the companies fired me and don’t blame them. Nobody wants to invite a shitstorm. I don't think this is why, or at least all of why. If I'm Wizards, and I'm promoting your art, organizing conventions and
974.
▲
by
Imnimo
5y ago
Maybe everyone's using this problem as their first screen: https://keithschwarz.com/interesting/code/?dir=find-duplicat...
975.
▲
by
Imnimo
5y ago
But ultimately, even if I trust someone else with my private key, if they lose it, it's irrecoverably lost forever. If Google loses my password, they just set me up with a new one, and I'm on my merry way.
976.
▲
by
Imnimo
5y ago
There are a lot of downsides to "Sign in with Google", but I am generally willing to accept them because I think I could recover my account if I lost my password. I'm not certain I could do so, because we've all read
977.
▲
by
Imnimo
5y ago
One might accuse the author of playing a game of his own, where his writing is driven by a desire to get clicks rather than a desire to communicate factually...
978.
▲
by
Imnimo
6y ago
I think this is the sort of thing that is very obvious if you are already comfortable with neural networks and the FFT. But if you're only comfortable with neural networks, and the FFT feels like arcane magic, this exercise might be v
979.
▲
by
Imnimo
6y ago
I was under the impression that adenovirus vaccines have only been been approved in the past year or two - a few for Ebola and the new ones for Covid. Are there others that have been around before that? It doesn't seem like it really
980.
▲
by
Imnimo
6y ago
I like to imagine this is the Jimmy Neutron origin story.
981.
▲
by
Imnimo
6y ago
Maybe they could save time by notifying the people who were -not- affected.
982.
▲
by
Imnimo
6y ago
Yeah but like, what if you find out that you can't distribute the remaining coins in a compliant way? Wouldn't that be something that should have been determined before all the work to integrate with Signal was done? It just fee
983.
▲
by
Imnimo
6y ago
Kinda seems like the sort of thing you would have figured out ahead of time...
984.
▲
by
Imnimo
6y ago
The use of the little plastic flippers to control the junctions reminds me a lot of the Dr. Nim toy: https://en.wikipedia.org/wiki/Dr._Nim It uses similar flippers to operate a marble-driven computer to play Nim.
985.
▲
by
Imnimo
6y ago
As a reader of papers on Arxiv, I get a lot of value from having author names and institutions on them, and having them all be by "a group of machine learners" would be a pain. I want to be able to find high quality work that
986.
▲
by
Imnimo
6y ago
Maybe he could be a Reddit admin...
987.
▲
by
Imnimo
6y ago
>All of this makes me believe we need more time for our brains to adapt to writing on glass, or technology needs to advance enough so that writing on tablets feels like writing in a notebook. This strikes me as implausible. Did the earl
988.
▲
by
Imnimo
6y ago
At this point just fill in that section of the canal, leave the ship there, and dig a new canal.
989.
▲
by
Imnimo
6y ago
If you've got a suspended license, better to let the car drive itself! /s
990.
▲
by
Imnimo
6y ago
Why would I buy a Grumpy Cat NFT when I could buy an entire real-life star on the star registry for only $50?
More ›