2 websites defaced with facts

Why not Fluxer?

Yeah unfortunately the new one doesn’t skip their neck out. I’d like way more pink

No worries, you are right. And even if typst was perfect you have to work with other authors or journals that won’t want to use it

I personally never had a case where Typst couldn’t do what I wanted and usually found something better in Typst instead. The only exception is to do curly snake-like rectangle borders, but it’s probably doable in Cetz.

As for hyperref, you are propably right, I would personally use a boolean flag for that and do two versions. Though the whole point of a pdf is to have the same look between it and the print, other features might fail depending on the viewer.

I think the extra spaces in math make it more readable, and it’s kinda the price to pay to not have to add backslashes everywhere. One instance when I found it weird was that frak("abc") had a different font than frak(a b c). So i just make a #let abc = $frak( a b c)$ rule.

There’s no going back from typst

I read a reddit thread about this.

Basically they are significantly safer because the review process is tedious and the PRs take ages to get reviewed. More over the read-only nature of the nix store make most of those techniques useless. You cannnot just take over packages the AUR way.

Moreover, if you use third party nix flakes, you are still safer because they are tied to a specific github repo, so if it gets forked by malicious actor you won’t get that update.

However you are still prone to upstream malware. That is nixpkgs probably won’t add malware but it could be there before packaging.

I use NixOS btw

The readme is gigantic with weird info like keybinds and project layout, especially the bizarre “what works” status section and the mention of number of unit tests.

do you have the source of this picture?

where is this from?

nixos comes with systemd

They beat any human on that knowledge benchmark, completely unrelated to your 40% “test”. Try to answer any of the example questions on the main page.

I don’t need a metaphor I know LLMs are hallucinating, lying, bullshitting. That doesn’t invalidate my point.

This is correct, I suppose you’re talking about the final softmax layer? When I said they are bad at determinism, I was talking about reasoning on deterministic rules not having deterministic output. For example, LLMs make logical deduction errors, calculation errors etc.

It does lie and hallucinate a lot, especially with biased context in the question (the bullshit part). The (biased) knowledge is hiding somewhere in its weights, it is just that it is sometimes quite hard to recover.

Your 40% depends a lot on how you ask the questions and the field of these questions. Humanity’s last exam is a morr obiective benchmark for measuring the wide knowledge of LLMs.

Yes and it’s frustrating

If you recognize there’s emergence, the “it’s just probability” take is misleading

The fact that it uses a non-trivial neural network. If it was simply a rate count of based on a corpus of how much time each word is followed by each it wouldn’t be stronger than keyboard word predictions. To make accurate suggestions requires emergence of primitive reasoning on the semantics of the tokens, LLM neural networks (transformers) can be analyzed to find subnetworks dedicated to modeling reality. It is still probability, but saying it’s just probability is not faithful

That it is not a calculator and is horrible at determinism is not debatable, however its (very biased) huge knowledge is its core feature

I have same wallpaper on nix :p