jbay808

7 ms·

31.
by jbay808 2y ago
The past decade has seen a huge number of problems widely and confidently believed to be "actual hard mode problems" turn out to be solvable by AI. This makes me skeptical that the problems today's experts think are hard aren
32.
Tool and Die Making (1953) [video](youtube.com)
1 points by jbay808 2y ago|0 comments
33.
by jbay808 2y ago
Good question. Temperature is a funny thing. When a system is far from equilibrium, the notion of temperature becomes a little unfamiliar. The thermodynamic temperature of the radiation is connected to the entropy of its power spectrum. LED
34.
by jbay808 2y ago
A photovoltaic cell also requires a temperature gradient. In the case of a solar panel, it's the high temperature of the sun that shifts its blackbody spectrum into a range where the cell can generate electricity. But if you were to he
35.
by jbay808 3y ago
Thanks; I'm working on products in this area and not especially satisfied by the parts available. Especially as Analog seems to be abandoning their micro-MPPT IC product line. My email is jacob at jbaylessconsulting dot ca. One of the
36.
by jbay808 3y ago
Do you have a spec sheet for your MPPT chip that you can share? The white paper and video presentation that you linked appear to be completely unrelated.
37.
by jbay808 3y ago
Great talk and very well presented.
38.
by jbay808 3y ago
No, but if a Canadian engineer put their stamp on work that was this sloppy, an apology wouldn't suffice. They'd be suspended at least, or likely even lose their license permanently . Even if the work was prepared by their subord
39.
by jbay808 3y ago
Absolutely not! The choice of fixed- vs floating-point for a controller should be based on domain knowledge of the problem, and the right choice may vary depending by operation. For example your PID's integrator needs to accumulate a r
40.
by jbay808 3y ago
It reminds me of the old story about the inventor in the late 1890s, who claimed to have invented a car that could cross the United States running on just a gallon of water. After failing to deliver a tangible product for years, investors p
41.
by jbay808 3y ago
Iron nitride is in many respects not better than neodymium, but certainly very promising nonetheless.
42.
by jbay808 3y ago
It is only difficult for a LLM to sort a list of numbers if the list is longer than half of the context window. (Source: I tested this myself[1]). The sorts are not error-free every time, but with sufficient training they become error-fre
43.
Barkhausen Effect(en.wikipedia.org)
5 points by jbay808 3y ago|0 comments
44.
by jbay808 3y ago
That was my concern as well...
45.
by jbay808 3y ago
As a Substack writer, I find it very distracting. Part of the reason is that when my eyes are looking at the body text, the motion happens in my upper peripheral vision. That's also where tab heading notifications are (eg. an incoming
46.
by jbay808 3y ago
> that's one preprint on arxiv, that makes a wild claim about a new concept that they acronymise as "RASP". It's not any kind of established terminology, nor is it anything but a claim. Would you change your mind for
47.
by jbay808 3y ago
I agree a neural network can certainly learn the conditional distributions that let it make that choice correctly every time. Once it has done so, then do you not have a sorting algorithm?
48.
by jbay808 3y ago
> In your comment above Oh, I see, good catch. I think that comment was a result of a botched edit; I do that sometimes. Too late to change it now. Sorry for the confusion! > Where does this use of "n-gram" come from? I mean
49.
by jbay808 3y ago
> The (sequential) distribution of digits amongst sorted numbers is tiny This is why 10^80 random lists gets reduced to only 10^36 sorted lists. However, 10^36 is still very large with respect to the size of the model.
50.
by jbay808 3y ago
> what do you think would happen if you gave your model an alphanumeric list to sort? Did you try that? The tokenizer would throw an exception, because it doesn't have any tokens to represent alphabetical characters. But you tell me
51.
by jbay808 3y ago
> it sure has not learned any "algorithm". At best it has learned to approximate a sorting algorithm. That's what statistical machine learning models do, they are function approximators; not program learners. Transformers
52.
by jbay808 3y ago
> If LLMs "learned algorithms", the best compression would be on the order of bytes. Yes. Except: (1) the model size is fixed during training, it would be impossible to obtain a bytes-sized result regardless of what it learns t
53.
by jbay808 3y ago
> Redness is not in the structure of those sentences. Sure; it's in the spectrum of reflected light. (Or perhaps, the retina's trichromal responsivity). But that physical concept can be meaningfully described by sentences. It
54.
by jbay808 3y ago
> you havent formulated any hypotheses that can be tested. You've just looked at the graph and "reckoned something" Let's not be so hasty. I think I do put it as clearly as possible. I'm comparing essentially you
55.
by jbay808 3y ago
> It is absolutely trivial to show Hyp2 is false To investigate precisely this question in a clear and unambiguous way, I trained an LLM from scratch to sort lists of numbers. It learned to sort them correctly, and the entropy is such th
56.
by jbay808 3y ago
On this note, is anyone aware of any two-dimensional colourmaps that are perceptually-uniform (or close to it), or perhaps periodic? I find I often have to hand-roll these myself.
57.
by jbay808 3y ago
They do not handle the rapid temperature differentials a home chef often encounters when they aren't careful enough. Especially when they are expecting the resilience of borosilicate glass.
58.
Thinking Like Transformers (2021) [pdf](arxiv.org)
112 points by jbay808 3y ago|20 comments
59.
by jbay808 3y ago
I strongly disagree with this. It only makes sense to use floats if you want time to have increasing precision the closer you get to 1970. From a symmetry group perspective, that makes no sense; the amount you care about the incremental pre
60.
by jbay808 3y ago
I won't take a hard stance on what counts as "reasoning", which I picked in the title for lack of a better summarizing word; I am open to alternatives. So if you think that making abstractions and implementing a sorting algor

More ›