khalic

3 ms·

31.
by khalic 18d ago
Clickbait title…
32.
by khalic 19d ago
That’s a lot of misconceptions for one paragraph…
33.
by khalic 19d ago
lol somebody didn’t hear the news about Kimi and GLN 5.3 cyber
34.
by khalic 22d ago
I can't wait for Qwen 27B on an ASIC
35.
by khalic 22d ago
Why is the headline trying to make it sound like it's new or a big deal? That's a terrible sensationalist website btw...
36.
by khalic 22d ago
Retaliation by Hegseth for not allowing Claude to be used for weapons systems.
37.
by khalic 22d ago
Fascinating read thank you
38.
by khalic 22d ago
"Nobody could have predicted this" It's getting really tiresome
39.
by khalic 22d ago
Indeed, but the fact that it moves the needle at all is enough to be skeptical of the anthropomorphic interpretation of the intermediate tokens. It's waaaay too early into the discovery process to take anything for granted, even more w
40.
by khalic 22d ago
Why does injecting random text increase the precision then?
41.
by khalic 22d ago
Scientific papers aren't popsci articles. Constraining the environment (like with the token choice) is a basic method to reduce confounding factors... The "toy model" part is flat out wrong, the models they use have _exactly_
42.
by khalic 22d ago
I believe it because of scientific papers on the matter. It doesn't "think" that's the point, it's generating text that describes a thought process, it's related but not the same thing.
43.
by khalic 22d ago
[2404.15758] Let's Think Dot by Dot: Hidden Computation in Transformer Language Models https://arxiv.org/abs/2404.15758
44.
by khalic 22d ago
I was under the impression that intermediate tokens (“chain of thought”) are _not_ a representation of a model’s logical path, with one study observing that you can replace intermediate tokens with single character chains and still get the
45.
by khalic 23d ago
Ah! verisign! Proving the world over and over what kind of scum they are
46.
by khalic 23d ago
check their status page...
47.
by khalic 23d ago
PSA you can switch to sonnet and compact as damage control
48.
by khalic 23d ago
Are they crumbs? Why aren’t you answering?
49.
by khalic 24d ago
Lol, are you saying the Erdos problems are a "crumb" that nobody bothered to empty? Are you doing a comedy routine?
50.
by khalic 25d ago
As explained in the article (...), they won't even have a parasitologist, and cyclospora specialists are rare. How about you read the article and then post? Both your questions would have been answered that way. If these are bad faith
51.
by khalic 25d ago
Why did you not include the rest of the sentence in your quote? > Although the research is still funded, insiders told Politico that not one of the scientists working on the project plans to relocate from Maryland to Iowa. Research isn&#
52.
by khalic 25d ago
Or they flat out want the old and vulnerable to die.
53.
Trump admin shelves Cyclospora research despite record-breaking outbreak(arstechnica.com)
10 points by khalic 25d ago|12 comments
54.
by khalic 1mo ago
Gonna have to wait a few days to see what the wizards of the HF community come up with…
55.
by khalic 1mo ago
thank you kind stranger
56.
by khalic 1mo ago
no i mean a replacement for nitter and xcancel there's still a lot of info you can only find on twitter
57.
by khalic 1mo ago
you won't get anything out of a whole book embedding anyway, even a structured page is too much
58.
by khalic 1mo ago
Any ideas for a decentralized version of these services? Maybe something tied into the fediverse?
59.
by khalic 1mo ago
So no, you didn't, the info comes from an app for parental surveillance. This is in no way a random sample that would be needed for reliable data AFP is subject to the same scrutiny as others, and definitely put out some terrible piece
60.
by khalic 1mo ago
And did you check where the information comes from?

More ›