cshimmin

15 ms·

121.
by cshimmin 4y ago
Understood! My point was that, while dimensional analysis almost always works (including for this problem), you have to make sure you are using the right dimensionful quantities. In this case that there are constraints here due to orbital m
122.
by cshimmin 4y ago
Gp is quoting one of Yogi Berra's oxymoronic quips. It's funny if you think about it for a second :)
123.
by cshimmin 4y ago
Hmmm, interesting analysis, but 12,000km is 3x wider than continental US. Based on the plot from ESA's projection, the track over the +-10m period is much smaller than France. I think the issue with your reasoning is that when you are
124.
by cshimmin 4y ago
Yes that's exactly right. You can see this information in the plot attached to the linked Twitter post. There is a range of times (indicated by color) and the corresponding projections for impact location. But note that meteors can be
125.
by cshimmin 4y ago
I disagree; consider the use of the term "video game AI", which historically at least has just been a bunch of _if_ statements chained together. This is totally valid, it's an example of AI without machine learning. The thing
126.
by cshimmin 4y ago
Haha I had to go and confirm, you're right (it's 35/55 for those who don't want to click a button 55 times). And here I was feeling pretty good about my 42/55...
127.
by cshimmin 4y ago
It actually could work pretty well. A Geiger counter is essentially a current gain device. Ionizing radiation passes through the chamber and excites free electrons which travel towards high voltage plates. The gain comes when the drifting p
128.
by cshimmin 4y ago
Sure but all that means is that 1g of radium will have the same activity as 0.2mg of polonium. Of course they will also have different decay spectra etc. We don't know how big the source was that the students lost, but I'm guessin
129.
by cshimmin 4y ago
When I get listicle spam I tack "site: reddit.com" onto my query and usually get pretty organic, opinionated results. But of course you might come up dry if what you're looking for is really obscure.
130.
by cshimmin 4y ago
People in this thread are pointing out difficulties and I'm getting an overall dismissive tone. As an ML researcher myself, I'm often skeptical and reserved about AI hype. That being said, the difficulties anticipated here are goi
131.
by cshimmin 4y ago
This is definitely not true in the US, at least in physical sciences. Typically your PhD advisor (who is obviously in the same organization and gets paid with the same funding source as the student) has almost entirely all of the say in a P
132.
by cshimmin 4y ago
@Dang / moderators: in case you are not already aware, in the sibling comment which is dead the username contains an inappropriate racial slur. Sorry I don't know how to better report this. Edit: I figured it out, sorry for the no
133.
by cshimmin 4y ago
Hmmm... June 11, 2020: MegaFace dataset is now decommissioned. University of Washington has ceased distributing the MegaFace dataset citing the challenge has concluded and that maintenance of their platform would be too burdensome.
134.
by cshimmin 4y ago
"I've heard of some people still having it as 5k iters even with an active account" I can confirm because I am one such person. Honestly thought that would come across from context but I'll edit to clarify.
135.
by cshimmin 4y ago
Yeah I was a daily user until about a month or two ago. Didn't even know about this setting, and I still haven't figured out how to change it in the mobile app. But mine is definitely set to 5k. Still in the process of changing al
136.
by cshimmin 4y ago
Can confirm :( Edit: because I was (until very recently) a daily active user and had niter=5k. Didn't even know about this setting until the hack.
137.
by cshimmin 4y ago
I'm pretty sure "our best minds" aren't the ones building the ape avatars. They are working on the fundamental advancements of AI itself. Once those tools are built, it doesn't take nearly as much "best brains&
138.
by cshimmin 4y ago
Fantastic writing, I would love to see a film made of this!
139.
by cshimmin 4y ago
Love secret Hitler, but I do regret the theme (someone has to actually RP as Hitler...) Was thinking of getting it for my sister who lives in Munich, but I wonder how people in Germany feel about this casual Nazi game? Luckily, one of the c
140.
by cshimmin 4y ago
Seems inspired by trump's playbook?
141.
by cshimmin 4y ago
No comment about Chesterton's Fence but I have heard (but never confirmed!) one reason for left-handed stigma is to do with hygiene. The idea is that before modern hygienic standards (e.g. sinks with soap in every bathroom/kitchen
142.
by cshimmin 4y ago
This seems to be to opposite of the example GP was seeking. Unless you mean to say that a great many people (both in Brazil and internationally) were already practicing Capoeira before it was legalized...
143.
by cshimmin 4y ago
Wow, from the 2015 article: [A journal reviewer of his famous paper says]: "you should be careful. This guy posted a paper once, and it was wrong. He never published it, but he didn’t take it down, either.’ ” The reader meant a paper t
144.
by cshimmin 4y ago
The answer to your question is right there in the sentence. They would like to draw attention to it.
145.
by cshimmin 4y ago
I think a really good example of this is adjective precedence. For example, any native speaker immediately knows "wooden nice green big house" is wrong, and they also immediately know that "nice big green wooden house" i
146.
by cshimmin 4y ago
And you are confusing aesthetics for ascetics :)
147.
by cshimmin 4y ago
This is noted in the article, "A direct causal link between musical training and general cognitive faculties, however, seems unlikely." The point of the article is that they've found a specific neurological relationship betwe
148.
by cshimmin 4y ago
FWIW the first line OCD meds are antidepressants (SSRIs). If you're not already on SSRI maybe you can look into switching and killing two birds with one drug.
149.
by cshimmin 5y ago
It's especially true for models that do some kind of weight sharing, which is very common (CNNs, RNNs, transformers, etc). For a concrete example, consider a layer from an image convolutional network, which maps from a 3-dim colorspace
150.
by cshimmin 5y ago
From the paper, they are using bfloat16, so I guess two bytes. But distributing and "packaging into an app" are not at all of practical interest for these kinds of models. You (a consumer) would interact via some API service, with

More ›