Hill climbing is a statistical method for finding maxima in a function. If your function's range is "does stuff well", then hill climbing is equivalent to finding improvements.
Also, Wikipedia's search is free to anyone on the internet.
3,238 karma · joined August 17, 2011
Also, Wikipedia's search is free to anyone on the internet.
I suspect the criticism isn't as absurdly narrow that ALL we have to do to get M4A is swap out the war funding. It's probably people pointing out that the govt wastes plenty of money on immoral and useless things, of which the Iran war is an example.
I'd bet only a small % of people praising TigerStyle actually disallow all runtime allocation in their own programs.
Do they, actually? I got the impression from the article that little revocation testing of any scale happens. If so, then even a limited test would be useful.
But what about you? Are you not doing https://xkcd.com/2501 ? Is your expectation reasonable on HN?
They'd be created with big red warning signs, installed into browsers and other places by anyone who wants to test, and then publicly revoked every month. Rinse, repeat.
It's a godsend for those looking to do what they want with old hardware. You can now pretend to be the authoritative domain for updates.
Roughly, my question is "Don't we see other post-collision galaxies that still have dark matter, and if so, why?"
If we have two identical scenarios with different outcomes under a theory, that doesn't bode well for the theory.
This kind of presumes dark matter has been definitively proven already, though.
From the outside, CDM may be a better theory than MOND, but it doesn't seem settled to me. I haven't heard anyone describe it as a "Law" yet.
Couldn't someone equally characterize the creation of dark matter as "adding and tweaking mass until you fit the data"?
It seems that having some galaxies without dark matter is also pretty bad for CDM. Why would CDM be absent in those cases? Without an explanation, isn't it bad for CDM (as well as MOND)?
Nobody wants a theory where exceptions can be hand-waved away as "Oh, that galaxy must not have DM" or "That galaxy has too much/too little/different balance/etc".
What's next, tobacco company bemoans anti-smoking ad restrictions?
Based on the timing, I'm deeply surprised it came out 2 years after the NES was released in the US, but I suspect it was way cheaper, and bought by some number of clueless parents.
E.g., a high-level auditory neuron might primarily fire in response to high-pitched voices, but if you hear someone say your name in a high-pitched voice at a party, you can bias that neuron to try more selectively fire in response to just their voice, "tuning out" similarly-pitched voices.
"It is very difficult to disagree with anyone of the caliber such as this man".
Nobody suggested you argue with Hickey to his face.
How attention affects biological neurons has little in common with the ML mechanism of attention.
Salience explains the "why", but I'm talking about the "how".
> Part of the reason was not knowing about generalized pose neurons
I have to ask, did they not look up how the human visual system works? The primary visual cortex has any number of early neurons that are spatially-tuned for various shapes and edges, and this is pretty basic knowledge.
In fairness, neural networks are more divergent from biology than a lot of people think, but it seems there's still low-hanging fruit to be grabbed with a minimal collaboration.
So instead of being a useful, if imperfect, analogy from an outside domain, trying to use ML attention to understand biological attention just results in confusion (like this article).
Neuroscientists aren't going to tell ML researchers they can't use "attention", but nor are they going to pay attention to them (pun intended). Neural networks are almost always too divergent from biology to shed light on real neurons.
As a former perceptual neuroscientist, I can assure you that ML "attention" has very little to do with biological attention.
"Spreading activation" would be a better analogy.
----
It's mostly harmless, though it occasionally results in head-scratchers from the biological POV. I once saw a ML paper talking about "retinal attention", which is funny, because the retina is one of the few places in the central nervous system where afawk, attention is impossible.
There's simply no efferent (brain->retina) non-motor nerves capable of biasing photoreceptor output; the retina is largely ballistic. (Eyeball motion is not attention.)
When you place someone on a pedestal, you make it hard to evaluate them critically. The remaining Clojure community has a problem with it.
Most of human brain activity is unconscious, yes, but p-zombies don't quite fit us imo, since they have NO conscious awareness by definition. I really feel it's meant to be a more present/absent distinction.
They're a good fit for Watts's aliens, though.
There's a lot more going on at synaptic clefts, so just knowing the number of synapses doesn't tell you enough to model anything. You still need to know which neurotransmitters are used, how much, second-order effects like G proteins, basal firing rates, distance to the axon hillock, the shape of the neuron's effect on potential decay, etc.
And that's all to predict whether one neuron will fire. You could get very different behavior between different two neuronal pairs having the same synaptic count.
"Magnetism is second nature to us electromagnetic chemists, so it's easy to forget that the average person probably only knows the formulas for one or two paramagnetic substances."
"And diamagnetic, of course."
"Of course."
Whereas "vampires" immediately call to mind Vlad, Dracula, Twilight, etc. and an entire literary tradition of the last century, which Blindsight has little to do with. There's the riffing on crucifixes with the right-angle seizures, but that's more cutesy than serious.
It's not the worst decision on Watts' part, but I think you're in the minority in liking it.
SKAT had some great headquarters, I'll say.