2-3-7-43-1807

8 ms·

31.
by 2-3-7-43-1807 2y ago
exactly ... But why is there so much sympathy with leaving the EU? I don't get it. /s
32.
by 2-3-7-43-1807 2y ago
Sabotaged by our incompetent and ignorant politicians for sure.
33.
by 2-3-7-43-1807 2y ago
And in Germany we are still debating if Syria is safe enough to send asylum seekers back there :D
34.
by 2-3-7-43-1807 2y ago
i don't think that is a valid argument.
35.
by 2-3-7-43-1807 2y ago
i think i misinterpreted your first sentence.
36.
by 2-3-7-43-1807 2y ago
i don't think that's what he meant and also don't think that is accurate to say they already have an understanding. i'm not even basing my criticism on the anthropomorphization but on the fact that there will be activati
37.
by 2-3-7-43-1807 2y ago
i don't think it's just about the training material. it's also about keeping track of the precise number of tokens. you'd have to have dedicated tokens for 1+1+1+1 and another one for 1+1+1+1+1 etc.
38.
by 2-3-7-43-1807 2y ago
what do you refer to by "other models" not belonging to the "SOTA ones"?
39.
by 2-3-7-43-1807 2y ago
> you need more than the model itself and a way to run it. You also need the data and the program that can be used to train it. the model reveals the architecture which is all you need to use/run/train it.
40.
by 2-3-7-43-1807 2y ago
> for instance "what is 2+2" or some numerical puzzles that needed algebraic thinking there is only one algebraic approach to solving something like 2+2 and that is counting! 2+2 = (((0 + 1) + 1) + 1) + 1). but llms are infamo
41.
by 2-3-7-43-1807 2y ago
> In a way this is obvious in hindsight, but it goes against ML engineers natural tendency when detecting a wrong answer: Teaching the model the right answer. But the answer space for LLMs is infinite and unbounded. So, no effort will be
42.
by 2-3-7-43-1807 2y ago
“Chop wood, carry water.” or rather “Chop wood, carry water and submit your blog to hacker news using a throwaway account.” ? ;)
43.
by 2-3-7-43-1807 2y ago
taking cocaine is a more hardcore crime than computer crime?
44.
by 2-3-7-43-1807 2y ago
wouldn't be surprised if warm showers actually lower the core temperature.
45.
by 2-3-7-43-1807 2y ago
if we look for agi, then not noticing the gorilla might be a good thing. i'm referring to the gorilla counting experiment.
46.
by 2-3-7-43-1807 2y ago
thanks for the link. will have a look at it. if you tokenize tiles and then feed those serially to an llm. i really wouldn't know why someone thinks that's a good idea. you lose all local spacial context not to mention global cont
47.
by 2-3-7-43-1807 2y ago
the llm isn't multimodal. an llm can only process textual tokens. what should those tokens be for pictures. the llm gets fed a textual representation of what was optically recognized by another process. that's my understanding.
48.
by 2-3-7-43-1807 2y ago
i dont understand. what have llms to do with ocr?
49.
by 2-3-7-43-1807 2y ago
pelosi is neck deep in insider training. and biden's pardons for his son and others are also "legal" at best. just to give two examples off the top of my head. if the law is broken - you have to break the law.
50.
by 2-3-7-43-1807 2y ago
the term "aware" qualifies your _opinion_ or feelings as a fact, which i refute. it is snarky, maybe, on the dark side of humour. but "offensive" - i don't think so.
51.
by 2-3-7-43-1807 2y ago
seems like someone is finally cleaning up that mess.
52.
by 2-3-7-43-1807 2y ago
the main narrative so far was that deepseek was cheaper and better than llms by openai.
53.
by 2-3-7-43-1807 2y ago
preparational activities for criminal goals are often decidedly legal. same for coups. you hollow out the defense of your opponents by legal means before you go for the guts.
54.
by 2-3-7-43-1807 2y ago
> Deepseek has just shown that you probably can't beat the big labs with these resources is that a new take? cause so far deepseek was considered as proof for small companies being able to compete with big players like openai ...
55.
by 2-3-7-43-1807 2y ago
> I think the idea was to wait for people mentioned in those documents to die, so as to not affect their privacy. you probably also think that mk ultra is a conspiracy theory.
56.
by 2-3-7-43-1807 2y ago
true ... afaiac it should be auto-flagged always. politics and also other stuff that belongs on reddit. i do like reddit, but i do like also that hn isn't like reddit. at least not yet. but it seems headed there.
57.
by 2-3-7-43-1807 2y ago
what a world where an adult doesn't even have to take responsibility for their own actions because of whatever excuse. sad state.
58.
by 2-3-7-43-1807 2y ago
maybe cause this is site is about technology and not politics?
59.
by 2-3-7-43-1807 2y ago
didn't know that's a meme
60.
by 2-3-7-43-1807 2y ago
did somebody confirm that this actually works? why didn't the author provide a copy pasteable text of the question in hex code? i call bullshit on this.

More ›