fspeech

7 ms·

1.
by fspeech 10d ago
Proofs are erasable. If you don't doubt it exists why do you care? Understanding is a side effect. Only people who want to understand the proof would need to care about it.
2.
by fspeech 11d ago
I don't think the truth of the theorem is ever in doubt so any attack would be silly. But the proof would enable tutorials like this: https://github.com/htzh/flt_for_human/blob/main/math/001-fre
3.
by fspeech 12d ago
The webpages are entirely generated without a binary build (a build from scratch is quite daunting as stated in the project readme) of Lean artifacts. See https://github.com/anthropics/fermats-last-theorem/blob&#x2
4.
by fspeech 12d ago
To really read the proof, clone the repo and drop the root index.html into your browser and enjoy. Due to the large amount of files in a directory Github won't serve the .lean files in Theorems/ beyond A. Github preview won't
5.
by fspeech 12d ago
Nice to get some resonance. And if I may, past --> taste; vision --> future. And to compensate for our poor memory, taste<=>value network while vision<=>policy network, borrowing from RL parlance.
6.
by fspeech 12d ago
It's 6B output tokens, as stated by the blog post.
7.
by fspeech 12d ago
First I have to say this is sooner than expected, even though I never doubted that this could be done. I am grateful that they dedicated resources to accomplish this. It is clear that agents are very good at discerning and holding onto very
8.
by fspeech 13d ago
Taste or vision? They are hill climbing and they can't see the other side. And sometimes that is because they don't live in your head and don't know what you want.
9.
by fspeech 1mo ago
Well it was made at least two years ago. It only just surfaced now.
10.
by fspeech 2mo ago
Linear layers use decays (like IIR filters) that naturally provide relative positions. Full attention layers can then be free to develop concepts that attend to each other regardless of distance.
11.
by fspeech 2mo ago
That's not true. First there is still a licensing and quota scheme on the US side for the H200s. Secondly China blocked them for use in inferencing. Thirdly Chinese companies don't want them for training because newer chips are mo
12.
by fspeech 2mo ago
Their kv cache is smaller so they can use less vram and also keep your prefix cached for longer. https://deepseek.ai/blog/deepseek-v4-compressed-attention
13.
by fspeech 5mo ago
Well, so far as the governments get VAT from the manufacturers, they are getting a return on their investments. They are more like mutualizing the companies than subsidizing them, that is the successful or mature ones are getting taxed to h
14.
by fspeech 6mo ago
Performance is generally limited by the process. Yield not so much. Assuming you can make it at a meaningful level at all, yield is generally a learning process.
15.
by fspeech 6mo ago
Yield is generally not an issue over time, at least not as big as someone outside of the industry would think, if you get enough chips to work in the first place. For high volume chips, fabs will tune their process specifically for your chi
16.
by fspeech 6mo ago
What do you mean? Yield is a function of the chip size and density as well as the process. Plus it's a commercial secret so your bet can't be adjudicated.
17.
by fspeech 7mo ago
Yes division is a poor example. It's a poor separation of concerns to try to wrap at this level without usage context. To see the point try to wrap overflows on arithmetic functions.
18.
Martial arts robots dazzle at 2026 Spring Festival Gala [video](youtube.com)
11 points by fspeech 7mo ago|3 comments
19.
by fspeech 8mo ago
Try ZhiHu(知乎).
20.
by fspeech 8mo ago
If you are just interested in a structural description (so-called netlist) the standard is EDIF.
21.
by fspeech 8mo ago
A lot of Chinese internet commentators are very ignorant of the reality in the US, but the Economist's riposte is weird too. For example how is the Chinese property malaise, which reflects an oversupply of housing and largely affordabi
22.
by fspeech 9mo ago
Price of a commodity metal can do whatever they want without causing a big problem. It is just a resource allocation signal. However if you base your currency value on it suddenly you are forcing debtors into bankruptcies if the value shoot
23.
by fspeech 9mo ago
Do you know what questions Newton was asking? https://en.wikipedia.org/wiki/Religious_views_of_Isaac_Newto... Being right is often hindsight and luck.
24.
by fspeech 9mo ago
Your point is right on. And additionally, why would an average Indian refuse the pay package to work in China? The top r&d guy at SMIC is from Taiwan after all. Liang got both Samsung and SMIC into the advanced nodes.
25.
by fspeech 9mo ago
Humans are also not rewarded for making pronouncements all the time. Experts actually have a reputation to maintain and are likely more reluctant to give opionions that they are not reasonably sure of. LLMs trained on typical written narrat
26.
by fspeech 9mo ago
I don't think people mind having bigger spaces but market is not clearing. In the US you have slums and bombed out building shells in prime urban locations as well. It is fascinating how human expectations work against each other.
27.
by fspeech 9mo ago
Cheap housing isn't the problem. The problems are people speculating on the appreciation of property, banking system depending on property value as loan collaterals and local governments depending on property sales for revenue generati
28.
DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning [pdf](github.com)
231 points by fspeech 10mo ago|50 comments
29.
by fspeech 10mo ago
I don't think the repair could be done. It's not about plugging a hole in space. It's about surviving reentry. They can't guarantee the integrity of the glass. Anyway to your point they could stock the kits in space on r
30.
by fspeech 10mo ago
It has a fixed capacity of how many different things it can pay close attention to. If it fails on a seemingly less important but easy to follow instruction it is an indicator that it has reached capacity. If the instruction seems irrelevan

More ›