nialv7

13 ms·

241.
by nialv7 2y ago
Or you could imagine an infinitely thin torus.
242.
by nialv7 2y ago
Yes, using those methods you can have a blurry idea of maybe this model is "smarter" than the other one, etc. But none of them may be used to anchor the claim that "intelligence is proportional to the log of compute", wh
243.
by nialv7 2y ago
This statement makes no sense because there is just no good measurement of intelligence for him to make a quantitative statement like that. What's he measuring intelligence with? MMLU? IQ? His gut feeling?
244.
by nialv7 2y ago
Plus level subscription has limits too, and Pro level costs 10x more - as long as Pro users don't use ChatGPT 10x more than Plus users on average, OpenAI can benefit. There's also the user retention factor.
245.
by nialv7 2y ago
Well yeah, if the llm can keep you engaged and talking, that'll make them a lot more money; compared to if you just use it as a information retrieval tool in which case you are likely to leave after getting what you are looking for.
246.
by nialv7 2y ago
Looks like more signal that the scaling "law" is indeed faltering.
247.
by nialv7 2y ago
How does this compare with Mamba? Intuitively it feels like this should have similar performance to Mamba, but the math is a lot simpler and easier to understand.
248.
by nialv7 2y ago
when's git-fsmonitor going to support linux?
249.
by nialv7 2y ago
i think they meant lower level.
250.
by nialv7 2y ago
just tried to play a round, the answer checker failed because it compares the results with `result == eval(expected)`, where `expected` is `"true"`, which is not a thing in Python.
251.
by nialv7 2y ago
It has been frustrating seeing so many people having the wrong opinion about AI. And no, that's not because I think one way (AI will take over the world! in more senses than one) or the other (AI is going to flop, it's a scam, etc
252.
by nialv7 2y ago
well, the problem is how far would you have to go? ok, you tell the AI to "not hack your opponent", what if they come up with a different cheating strategy? if you just say "don't cheat", what if they twist the mean
253.
by nialv7 2y ago
Honestly I think this is drop as in drop shipping.
254.
by nialv7 2y ago
why not wipe the device first before physically destroy it?
255.
by nialv7 2y ago
Oops. Thanks for pointing out.
256.
by nialv7 2y ago
There are more structures in non-computable numbers. As an example, a non-computable number could be "definable", i.e. you can describe it. For example, a real number the nth digit in the binary representation of which equals to w
257.
by nialv7 2y ago
*even if you pay up.
258.
by nialv7 2y ago
> We know RLHF and alignment degrades model quality. I feel you can't make statements like this without giving some sources. IIUC, without RLHF/alignment, the model won't even be able to chat with you, it would just be a d
259.
by nialv7 2y ago
Any project that has been successful for long enough will eventually become a legacy project. So I guess it's not wrong.
260.
by nialv7 2y ago
Just the other day I was trying to create a new Google account, and was surprised by how much PII it was asking from me. I have a handful of Google accounts, the newest one created I think almost of a decade ago? I don't remember I nee
261.
by nialv7 2y ago
jellyfin is open source, that's enough reason for me.
262.
by nialv7 2y ago
Yes, AI is going to solve all of our problems. /s
263.
by nialv7 2y ago
> I am sure by 2100 we will hit 4C I am kind of expecting 4C by 2050... given the current AI boom. Most protesting is not going to be effective (in the sense that they won't get people's attention). Effective protests will be h
264.
by nialv7 2y ago
i have to admit i tend to get drawn in by dramas like this, it is a fault of mine, but it is what it is. so anyway, i dug into the other party involved in this, i.e. the accused, and it looks like there is more to this story than what Lina
265.
by nialv7 2y ago
Oops, you are right. I think because the other direction of libffi - ffi_closure - has a jitted trampoline, I mistakenly thought both directions are jitted. Thanks for the correction.
266.
by nialv7 2y ago
I am surprised many don't know how libffi works. Yes, it does generate native machine code to handle your call. Look it up. Yes it's probably worse than doing the jit in Ruby interpreter, since there you can also inline the type c
267.
by nialv7 2y ago
isn't this exactly what libffi does?
268.
by nialv7 2y ago
Hey, they are using Mamba! Happy to see Mamba is used at least somewhere :/
269.
by nialv7 2y ago
I am skeptical. Intuitively I don't see what self-play achieves beyond straight RL. Have the authors done a comparison with the performance they can get by RL finetuning a single model by itself? Also this style of tasks is prone to ov
270.
by nialv7 2y ago
this model is specifically trained for solving math problems, so ask it some math questions?

More ›