petters

8 ms·

1.
by petters 25d ago
You likely just got lucky.
2.
by petters 1mo ago
I have lost commands in the Bash history as well and I don't know why. Fortunately, I store all commands executed (+ time, return value etc.) in an sqlite database so I run a program from time to time to import that into the Bash histo
3.
by petters 1mo ago
Only 1000x overhead would make some image classification tasks go from 1ms to 1s. That’s viable for some applications!
4.
by petters 1mo ago
These are also very cool: https://antfu.me/posts/ai-qrcode Diffusion image generation constrained to also be a valid QR code
5.
by petters 2mo ago
Why does figure 1 use the same color and style for two different graphs?
6.
by petters 2mo ago
> Since connection poolers reuse connections between clients, the connection state of one client “leaks” into the connection state of another. Wow this is very bad. This actually happens in typical Postgres setups?
7.
by petters 2mo ago
5B revenue is not the same as 5B in the bank collecting interest.
8.
by petters 3mo ago
> a hand tuned input loop written in C that takes <<1ms Yes, I would certainly expect much less than 1ms. Perhaps 1µs should be the goal?
9.
by petters 4mo ago
> If we give an LLM a prompt that reads “The following is a conversation between Julius Caesar and Genghis Khan,” it will generate a coherent dialogue between the two historical figures. But no matter how detailed the responses are, no m
10.
by petters 4mo ago
It also means you'll have to onboard a lot of new employees all the time. That sounds exhausting.
11.
by petters 4mo ago
Wikipedia is amazing. The Swedish articles are even longer.
12.
by petters 5mo ago
Karpathy replied to IS with ”agree” at the time
13.
by petters 5mo ago
Many dismiss Dawkins here but Ilya Sutskever wrote in 2022: “it may be that today's large neural networks are slightly conscious.”
14.
by petters 5mo ago
We have a very good idea of all math behind chemistry. But the equations are very difficult to solve.
15.
by petters 5mo ago
He is much better at building hardware than he is writing software.
16.
by petters 5mo ago
That's a good idea and it exists: https://www.johndcook.com/blog/2026/04/18/qlora/ It seems quite wastful to have two zeros when you only have 4 bits it total
17.
by petters 5mo ago
You could add a feature where it will compute the global optimum of any function of a small number of variables. Branch and bound with interval arithmetic works well for a small number of variables. Disjoint unions of intervals seems like a
18.
by petters 5mo ago
Yes, that blog post could have been much shorter….
19.
by petters 5mo ago
They have found a large number in OpenSSl
20.
by petters 5mo ago
Dungeon Crawler Carl is not science fiction. At least I would not recommend it to someone looking for science fiction with “interesting ideas.” It’s a comedy about an RPG with magic. But if that’s what you’re looking for, it’s pretty good
21.
by petters 6mo ago
> the GPU is limited by the Thunderbolt port Not everything is limited by the transfer speed to/from the GPU. LLM inference, for example.
22.
by petters 6mo ago
> On a 34" ultrawide monitor, it was too easy to put YouTube running on the left side, and whatever else on the right. Yes, if you were doing that, almost any change to your environment that stops that will be good. I don't thi
23.
by petters 6mo ago
That challenge was pretty stupid. I could read the question and I’m not even a native speaker. We can of course easily come up with much better challenges
24.
by petters 7mo ago
As a TA, I've seen adults try to pass initial college calculus many times (and failing - you were allowed to try several times) with enormous effort. It's not a small multiplier And this was still people selected from the small su
25.
by petters 7mo ago
I think that bias is not due to the proportion of books and more due to how they are fine-tuned after the pretraining.
26.
by petters 7mo ago
We identify the real number 2 with the rational number 2 with the integer 2 with the natural number 2. It does not seem so strange to also identify the complex number 2 with those.
27.
by petters 7mo ago
Going from 220 to 450 ms would be a disaster in my project. It has many thousands of files. Recompilation of almost everything happens from time to time. If those made-up numbers were true, they would be very significant and an argument in
28.
by petters 7mo ago
That's fair. I'm unable to provide more information though so we'll have to disagree.
29.
by petters 7mo ago
> C++ doesn't take longer to compile if you don't abuse templates. Surprisingly, this is not true. I've written a C++ file only to realize at the end that I did not use any C++ features. Renaming the file to .c halved the
30.
by petters 7mo ago
> We build Claude with Claude. Yes and it shows. Gemini CLI often hangs and enters infinite loops. I bet the engineers at Google use something else internally.

More ›