fjkdlsjflkds

12 ms·

61.
by fjkdlsjflkds 3y ago
You are probably thinking of forward-mode autodiff (which is more useful when the dimensionality of the output of your function is comparatively large), which is different from backward-mode autodiff (which is more useful when the dimension
62.
by fjkdlsjflkds 3y ago
Yes, but it does not necessarily have those two connotations in other languages (which is the issue here). In Portuguese, if you want to say "fresh water" (as in, "low-salinity water"), you use the term "sweet wat
63.
by fjkdlsjflkds 3y ago
> But libgen doesn't have "Good night, Gorilla". https://annas-archive.org/search?q=good+night+gorilla
64.
by fjkdlsjflkds 3y ago
I guess op meant "vec(ABC)=kron(B.T,A)vec(C)", and my attempt at explaining it would be: If you take the result of transforming the columns vectors in the C matrix by AB and vectorize it you get the same as vectorizing first C and
65.
by fjkdlsjflkds 3y ago
> markets can't go below zero and also can't jump by more than a certain amount per day either It depends on what you are looking at, and how you are modelling it, isn't it? For example: the value of log(asset_value_today&
66.
by fjkdlsjflkds 3y ago
No mentions of https://arxiv-sanity-lite.com/ ?
67.
by fjkdlsjflkds 3y ago
This is not entirely correct, from what I understand. The source of the "anomalous token phenomenon" is that those texts were included while training the tokenizer but not included when training the models. It is not clear the
68.
by fjkdlsjflkds 4y ago
If you continue reading that Wikipedia article, you'll reach this point: > A second-order Markov chain can be introduced by considering the current state and also the previous state, as indicated in the second table. i.e., a higher-
69.
by fjkdlsjflkds 4y ago
The burning sensation of "spicy" things is not mediated by "taste buds". For more information: https://en.wikipedia.org/wiki/Chemesthesis
70.
by fjkdlsjflkds 4y ago
> Integration and differentiation have interesting audio effects An integrator is just a low-pass filter and a differentiator is just a high-pass filter, so you'll most likely get "boring" audio effects ;)
71.
by fjkdlsjflkds 4y ago
"ability to evaluate and make nearly instant judgments in unique circumstances without prior training" It could be argued that there are already ML systems that do what you describe (i.e. make instant effective judgements/dec
72.
by fjkdlsjflkds 4y ago
Worse. The mechanism through which "substance addictions" and "behavioral addictions" happen (and are maintained) pretty much involves the same neurotransmitter paths: "Preclinical evidence has demonstrated that mar
73.
by fjkdlsjflkds 4y ago
Thanks for the correction. In that case, Google Translate's attempt at parsing that word completely failed: it seems to interpret it as "niemi-arytmiczny" or "niemia-arytmiczny", rather than as "nie-miarytmiczn
74.
by fjkdlsjflkds 4y ago
The reason why it can "make up words" is because it does not use "words", but "tokens" (which can be smaller or larger than a single word). In this specific case, it probably understands that the token "ni
75.
by fjkdlsjflkds 4y ago
No. You got the right impression. It is indeed doing "next token prediction" in an autoregressive way, over and over again. The best source would be the GPT-3 paper itself: https://paperswithcode.com/method/gp
76.
by fjkdlsjflkds 4y ago
There are (several) central limit theorems for dependent variables, but you often have to assume other things as well (e.g., stationarity, bounded third moment, and/or limited-range correlations) for it to work. Example: https:/&
77.
by fjkdlsjflkds 4y ago
"Central Limit Theorem" says that "adding (bounded variance) stuff together makes it converge to a Gaussian" (roughly). On the other hand, when you are adding random variable together, you are actually convolving their d
78.
by fjkdlsjflkds 4y ago
> Plagiarism is the *fraudulent representation* of another person's language, thoughts, ideas, or expressions *as one's own original work*. Source: https://en.wikipedia.org/wiki/Plagiarism The same way tha
79.
by fjkdlsjflkds 4y ago
> When people ask GPT-3 to write a rhyming poem, you see plenty of examples of GPT-3 poems starting with "There once was a cat named Pat..." This is an extremely common first line of a limerick, found anywhere from here[1] to h
80.
by fjkdlsjflkds 4y ago
Are you sure it is impossible? https://arxiv-sanity-lite.com/?q=diffusion+image+editing&ran...
81.
by fjkdlsjflkds 4y ago
"we know it's plagiarizing and combining other, third-party, materials half the time" does not seem like an accurate description of what a model like GPT3 does either, though, since: 1) No explicit content (that could be plag
82.
by fjkdlsjflkds 4y ago
This is actually a thing, yes (and it does increase robustness to adversarial examples). This is particularly useful if you apply (random) low-pass filtering (or denoising, or DCT-based compression, or anything that messes up high-frequency
83.
by fjkdlsjflkds 4y ago
So, would you claim that: "There is no probability whether an *electron* is on the path towards earth. It either is or it isn't."? I guess you must know something that Heisenberg didn't. Good luck defining things such as
84.
by fjkdlsjflkds 4y ago
> In reality it either is or it isn't. If you have good explanations like we do *in physics you don't need probability*. So... are you saying Statistical Mechanics (to give an example) is not part of Physics? In real life, the
85.
by fjkdlsjflkds 4y ago
I think I got your question right (but perhaps I was not totally clear in my answer). In a nutshell: sure, there are lots of such possibly useful models that use trees (you can even come up with new ones yourself). Random forests or GLM reg
86.
by fjkdlsjflkds 4y ago
> Is there some Bayesian version of decision trees where you model each decision in the tree as a beta distribution or something? Then you could simply choose the leaf or path with the highest confidence. Sure... there's many things
87.
by fjkdlsjflkds 4y ago
Nope. Decision trees typically output binary (or multinomial) values directly. Sure, you could tune the node thresholds (or optimize them during tree construction), to vary the trade-off between specificity and selectivity, but this may not
88.
by fjkdlsjflkds 4y ago
If you think that anything written in 4pt font on a website can turn "fraud" into "not fraud", I got some sweet bridgecoin I could sell you.
89.
by fjkdlsjflkds 4y ago
I stand corrected, then :) cheers.
90.
by fjkdlsjflkds 4y ago
This is not really "stable-diffusion based image compression", since it only uses the VAE part of "stable diffusion", and not the denoising UNet. Technically, this is simply "VAE-based image compression" (that

More ›