Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
lossolo
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
lossolo
8d ago
Yeah, I'm just curious about the setting. It's just weird to me that this wasn't disclosed by either party while the accusations were being made, that's all. Even if it was used in the training data, I don't believe
2.
▲
by
lossolo
8d ago
> If they opted out of training, then we definitely did not train on them. Can't you guys just check their account settings so the public knows what was set? EDIT: Why was this downvoted? I'm genuinely asking because I have no
3.
▲
by
lossolo
8d ago
Not entirely, it's just a late stage of the overall training process. It's an early checkpoint in post training (you can use the model at different stages of training), so it will probably become even stronger with more post train
4.
▲
by
lossolo
13d ago
Yeah, basically they are using more computation to explore the solution space before producing the final answer.
5.
▲
by
lossolo
13d ago
It was working for me too. sample_size++;
6.
▲
by
lossolo
15d ago
> a larger read is generally as fast as multiple smaller one on modern hardware. Not always if by modern you mean NVMe drives. One synchronous preadv() for 256 KiB gives the kernel/device one big request but 16 independent asynchron
7.
▲
by
lossolo
16d ago
A few days ago, my mother asked me to fix her YouTube because she kept getting something telling her to update. I looked at YouTube on her iPhone and saw an autoplaying video ad that displayed an animated fake system style pop-up button, de
8.
▲
by
lossolo
21d ago
How can you do so badly with Git when its architecture is basically so friendly to partitioning and horizontal scaling?
9.
▲
by
lossolo
23d ago
It's like saying that horse can understand physics, because he knows when to jump using his intuition.
10.
▲
by
lossolo
24d ago
> At some point chatting with Fable inevitably leads to it thinking about the security related aspects, tripping the safeguards. It happens to me all the time with things that have nothing to do with security, Fable spawns a subagent th
11.
▲
by
lossolo
25d ago
> Gross debt to GDP ratio in the US has fallen from 130% in 2021 to under 125% now. Most of that was inflation, so it hurts U.S. citizens. I mean, you could reduce it even further pretty easily, but I'm not sure you would want to pa
12.
▲
"Two 2030 AMD racks are expected to deliver same compute as 570 racks in 2024"
(newsroom.amd.com)
12 points
by
lossolo
28d ago
|
3 comments
13.
▲
by
lossolo
29d ago
Khan (ICC prosecutor) said Lindsey Graham (US senator) told him during a May 2024 call: "This court is for Africa and thugs like Putin. It is not for democracies like Israel and the United States of America."
14.
▲
by
lossolo
29d ago
If they remove that extra 50%, I'm moving my $200 to Codex. Good thing we will know tomorrow, just a day before renewal. "Your subscription will auto renew on Aug 20, 2026." But I doubt they will remove it, just like they did
15.
▲
by
lossolo
1mo ago
Why weren't the points merged again from the "dupe" thread that had 289 points? https://news.ycombinator.com/item?id=49299684 What a weird mechanism. If someone is judging a thread/topic/event impac
16.
▲
by
lossolo
1mo ago
> but this is still just GLM 5.2 with post-training magic. So exactly the same as Opus 5 and GPT 5.6 Sol. It's all "post-training magic".
17.
▲
by
lossolo
1mo ago
This is basically the answer, they generate A LOT of synthetic task rollouts in parallel, then use RL on the resulting reward signals to improve the model. Add scale to this and you have a Fable class model.
18.
▲
by
lossolo
1mo ago
Or you can also just place one or more tables on tmpfs, we are doing that in production.
19.
▲
by
lossolo
1mo ago
> I'm wondering what people expect these companies to do? Human moderators? Sure, but Meta products have approximately half the worlds population using them In this case, we are talking about ad images. 1000 moderators x 5 second pe
20.
▲
by
lossolo
1mo ago
What about it? TSMC now has fabs on U.S. soil. Intel and Samsung also have fabs and are working on processes that compete with TSMC's, even if their yields may be somewhat lower. Washington will have to decide whether the U.S. is willi
21.
▲
by
lossolo
2mo ago
> There are many low-hanging fruit that can be optimized to increase that revenue by 10x. The $15 billion figure is for FIFA's ENTIRE 2023-2026 revenue cycle, so not just this World Cup. But the obvious "low-hanging fruit"
22.
▲
by
lossolo
2mo ago
> and many others Trump and Kushners "Joshua Kushner is now the kingpin in driving the sale of the World Cup to investors and family ties are tight around the project"[1] 1. https://www.theguardian.com/football&
23.
▲
by
lossolo
2mo ago
> It's not difficult to understand the folly of sharing data and control with your biggest adversary. It's so basic, but people chronically keep tripping on it, intentionally or not. Like those self driving, remotely controllab
24.
▲
by
lossolo
2mo ago
So it's more like your personal definition, not the normal meaning of the word. Ok, because "janky" is informal slang also meaning poor quality, unreliable etc. I understand that you're using janky as shorthand for not O
25.
▲
by
lossolo
2mo ago
> Kudos to Moonshot for making these weights available for download. Lets not fool ourselves and claim these are "open source" by any understanding of the concept though, there are usage restrictions (even if you download them)
26.
▲
by
lossolo
2mo ago
Yeah, besides that, the only big other open source US model that was worth looking at was Inkling (975B params), Jul 15, 2026. https://thinkingmachines.ai/news/introducing-inkling/
27.
▲
by
lossolo
2mo ago
In what way was the license "janky"? Moonshot isn't a trillion dollar US behemoth. If these terms allow it to release near frontier models with open weights, so startups and anyone with enough hardware can use them, while cha
28.
▲
by
lossolo
2mo ago
It's showing you're switched to 4.8, i just hit that while doing security research.
29.
▲
by
lossolo
2mo ago
> Constant bugs, dropped sessions, issues switching models, http errors. It's becoming ridiculous And memory leaks.
30.
▲
by
lossolo
2mo ago
> Sure, these are my sources. World Bank (2023)... My WTO comparison used the much broader "medical goods" category, so it was not an apples to apples. > My source was Citeline's 2026 annual review at https://
More ›