pizza

14 ms·

211.
by pizza 2y ago
The more you spend on arxiv, the more you save on the gpus Jensen told you you would save more on if you were to spend more on gpus
212.
by pizza 2y ago
Luckily we have Shampoo, SOAP, Modula, Schedule-free variants, and many more these days being researched! I am very very excited by the heavyball library in particular
213.
by pizza 2y ago
Use a different part of your body, and offload the ‘write log’ to an external medium/buffer. Eg if you just constantly type at a computer screen and have recurring moments of waiting for your mind to recall what you are attempting to d
214.
by pizza 2y ago
You have to start trusting yourself and valuing what you find interesting more. One caveat to my advice though is if you have dependents in which case I'm not the right person to ask hah
215.
Side-Channel Attacks on Optane Persistent Memory(usenix.org)
1 points by pizza 2y ago|0 comments
216.
I've got 10M songs in my pocket. Now what? (2012)(slideshare.net)
3 points by pizza 2y ago|1 comments
217.
by pizza 2y ago
The value of the <EOS> token has gone up since then.
218.
by pizza 2y ago
I don’t think suggesting that his quote would imply his slaves would be justified in violating their own enslavement is any kind of presentism.
219.
by pizza 2y ago
Maybe something like dividends for all citizens? Either way, atoms will still matter where no number of bits will fix your problem. It's also possible that AI will introduce unimaginably profitable ways for non-technical people to give
220.
The most viewed articles of 2024(en.wikipedia.org)
33 points by pizza 2y ago|16 comments
221.
by pizza 2y ago
To explore behaviors of learning algorithms
222.
by pizza 2y ago
Pretty neat!
223.
by pizza 2y ago
Fair point. Actually probably the best part about having beaucoup bucks like Open AI is being able to chase down all the manifold little ‘last-mile’ imperfections with an army of many different research teams.
224.
by pizza 2y ago
Exactly, it trivializes the actual structure of the real world problem setup. It’s ML’s spherical cow.
225.
by pizza 2y ago
Holy moly.. even just the Llama 8B model trained on R1 outputs (DeepSeek-R1-Distill-Llama-8B), according to these benchmarks, is stronger than Claude 3.5 Sonnet (except on GPQA). While that says nothing about how it will handle your particu
226.
by pizza 2y ago
I’m still trying to figure it out myself too. I follow Victor Taelin on Twitter as he seems to be the most involved in the largest interaction combinator works going on. I suppose you could also try to explore the earliest literature on int
227.
by pizza 2y ago
Depends on where you're starting from 1. Absolute zero? Find the top rated short book in the language you're interested in. If you don't know which language, since you eventually want to get into AI, Python probably would be
228.
Testing Is Orthogonal to Correctness(blog.vmchale.com)
1 points by pizza 2y ago|0 comments
229.
by pizza 2y ago
> Also, expect to see that Facebook is partnering with TikTok on Monday morning. The head of the bill banning TikTok just invested 100 million in Meta... so I imagine there will be a followup announcement how Trump brokered some deal to
230.
by pizza 2y ago
Why isn’t the latter the situation right now?
231.
by pizza 2y ago
I’d rather have both data privacy protection through both laws and technological security AND a popular platform run by an adversary, than have neither privacy protections and only platforms that conceal their beneficiaries.
232.
by pizza 2y ago
And yet, the overall path of unconcealment of science and technological understanding definitely traces a line that goes from the Wright brothers to Vostok 1. There is no reason to think a person from the time of the Wright brothers would
233.
by pizza 2y ago
> For example, do we expect an AGI to be able to meditate, or to mentally introspect itself like a human, or to describe its inner qualia, in order to constitute an AGI? Do you mind sharing the kinds of descriptive criteria for these beh
234.
by pizza 2y ago
What about if the human brain were so complex that we could be complex enough to understand it?
235.
by pizza 2y ago
Maximum likelihood training tinges, nay, corrupts, everything it touches. That’s before you pull apart the variously-typed maximum likelihood training processes that the artifice underwent.. Your model attempts to give you a reasonably maxi
236.
Interaction Processing Units(nilscrm.github.io)
2 points by pizza 2y ago|2 comments
237.
Digital and Tunable Genetically Encoded Tension Sensors(onlinelibrary.wiley.com)
1 points by pizza 2y ago|0 comments
238.
by pizza 2y ago
A bit presumptuous to go into a rant that boils down to calling Jeremy an unskilled idiot. Jeremy has continuously used his programming experience and common sense to make more of an impact via his open source libraries, his courses on deep
239.
by pizza 2y ago
I guess it's not clear to me why it's even interesting to talk about their LinkedIn or their PhD in the first place? It's not like not having a PhD will make the work any more true or not. Wouldn't it be more interesting
240.
by pizza 2y ago
Why bother doing a witch hunt and leaving out that they did Stats at Yale..

More ›