babel_

15 ms·

31.
by babel_ 3y ago
Swapping variables around is also really handy for patterns like "alter list a until stable", especially with the walrus := operator allowing you to avoid declaring b outside of the while loop, keeping the intent clearer.
32.
by babel_ 3y ago
Must admit I wasn't quite sure what the point of the first one was. 0xBEAF is literally just 48815, as is 0x_B_E_A_F, as the numeric literals now ignore interleaved "_", and you just format it as a float after-the-fact as nor
33.
by babel_ 3y ago
> It's not fair to compare it to mature, well-developed stuff Really? So rather than try to compare, contrast, and course-correct a project in its early stages by understanding the priors and alternatives, we should only do retrospe
34.
by babel_ 3y ago
I saw this presented at CHI, and I think it's a ingenious and elegant sensory illusion that could make jumping in certain VR games actually feel immersive with haptic feedback and the ability to more closely match the height the avatar
35.
by babel_ 4y ago
They probably got them. The problem is that it's difficult to repeat, thanks to temperature, meaning users will get a random spread of outcomes. Today, someone got a legal game. Tomorrow, someone might get a grandmaster level game. But
36.
by babel_ 4y ago
Ironic.
37.
by babel_ 4y ago
I believe you're mixing your metaphors, as it were. A paintbrush is not like a neural network, nor is it a function. You may describe a paintbrush in a model, and that may be as a mathematical function, but a paintbrush it not itself a
38.
by babel_ 4y ago
Outside the scope of the paper, but sure, it's a valid and interesting question. If a model were to produce a pre-existing image not in its dataset, then we can strongly suspect that two behaviours have occurred (for the sake of establ
39.
by babel_ 4y ago
I mentioned isomorphism in reference to representation, which was kinda the point. And, not to belabour the point, but a function F that has an isomorphism to the real numbers means that F does indeed encode all numbers (and, when defined a
40.
by babel_ 4y ago
The argument here is more that this attack serves to just "find" these, and while it uses significant compute to do so, it's also not using many of the complex and more efficient extraction approaches that have been used on G
41.
by babel_ 4y ago
The tweet/paper co-author posted the paper ( https://arxiv.org/abs/2301.13188 ) yesterday on HN ( https://news.ycombinator.com/item?id=34596187 ) and ironically the top comment there is referencing to
42.
by babel_ 4y ago
Because the dual of a structure is clearly not another representation of the structure. Isomorphism means nothing. /s Your intuition is correct, fitting for the data will lead to this arising. However, you misunderstand the interpretat
43.
by babel_ 4y ago
You can lower the bounds significantly. Images do not require comparison to all others, and the presented approach to detecting and deduplicating images in the paper is easily adapted to be near linear (trading for storage or cleverness). P
44.
by babel_ 4y ago
The paper itself actually goes to show that this is insufficient (but should still be done, they went from regenerating 1 in 819 to 1 in 1063 images in a small corpus litmus test), and that diffusion is weaker to image extraction attacks th
45.
by babel_ 4y ago
The "black magic" of binaryFloat is a known type pun to a sign-magnitude integer, so because it searched for positive values it always worked, albeit it took a few steps to converge, however for negative values it is not correctly
46.
by babel_ 4y ago
Reaching for -O3 as the only option feels like a problem with the methodology. Is there the chance for -O2 or -Os as supplementary graphs? Or stacked / marked onto the existing graphs? -O3 is optimising for speed at a space-cost, after
47.
by babel_ 4y ago
His later statement that plausible states are only thus if they "evolve backward without generating contradictions" which is a truism if I ever heard one. You mean plausible states come from a prior plausible state, and can't
48.
by babel_ 4y ago
I don't know whether there was some clever construct to the thought experiment that was lost by the time of printing, or whether it's just a bizarre case of "I added in something that violates conservation of energy and looki
49.
by babel_ 4y ago
As the kids say: podders.
50.
by babel_ 4y ago
$1599 for 4090, $1199 for 4080 16GB, $899 for 4080 12GB (aka 4070). See Nvidia's News page for the announcement. Edit: Apologies, got the 4080 numbers wrong, accidentally used GBP, corrected above. For the curious, the 4080 16GB is £12
51.
by babel_ 4y ago
The $0.01 savings is for the maker, and not for the buyer. For a typical cable, it's probably around $0.30 to $0.50 to make, and they tend to sell between $5 and $15, depending on the spec. For simplicity's sake, we'll assume
52.
by babel_ 5y ago
Clearly history wasn't something it paid attention to in class. "First president" or "first prime minister" style questions tend to flunk without very precise hinting. Very enthusiastic about high quality models tha
53.
by babel_ 5y ago
I'd say that clearly there's a design choice in whether or not to have it, so Rust choosing to have these whereas Go choose not to is clearly something worth crediting as "probably good design". While I think there'
54.
by babel_ 5y ago
Subsidies simply offset the cost, they don't magically make those portions of it disappear. If you were to remove them, then by some estimates the price would go up anywhere between 30% and 180% for meats, because that's how much
55.
by babel_ 5y ago
Well the long and short of that is the infrastructure and space usage of livestock etc. aren't ignored, they're part and parcel of what the cost of bringing meat to your plate is. Can't say what exact breakdowns are, likely t
56.
by babel_ 5y ago
So indeed the simpler techniques are the fastest by walltime on the solvable problems; typically shown as N=16 through N=20, or up to N=22 or N=24 if you have sufficient parallelisation available. As you mention, it's not convincing th
57.
by babel_ 5y ago
This work only established new (remarkably exact) bounds, it was not, as the quanta article poorly describes, answering the problem in the exact sense. Their "all but solved it" is a little disingenous, as indeed it balloons quick
58.
by babel_ 5y ago
Having done academic research into N-Queens, I'll say that this is result is a remarkable step forward and, once vetted, is generations' worth of advancement in establishing the upper and lower bounds of the problem with remarkabl
59.
by babel_ 5y ago
I would say that the key benefit for this isn't really in making mature codebases nicer (though it can), but rather for rapidly-developing codebases, especially when doing prototyping or iterative development. The structured nature and
60.
by babel_ 6y ago
Interesting follow up reading: "Relaxing the Constraints on Predictive Coding Models" ( https://arxiv.org/abs/2010.01047 ), from the same authors. Looks at ways to remove neurological implausibility from PCM an

More ›