Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How does GPT obtain its ability? Tracing emergent abilities of language models

yaofu.notion.site

Post reply on HN
Print view View original

161–170 of 205 posts

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#161

Post by sandgiant »

chriswarbo wrote: post #149
benreesman wrote: post #147

I have yet to see an output from a big language model that doesn’t just look like P(text|internet). I understand that it’s very easy to ascribe all kinds of qualities to these things, but when the corpus is the Internet, the log likelihood of it sounding like a person is not so different from the corpus sounding like a person. These things are impressive enough without any magical thinking.

> I have yet to see an output from a big language model that doesn’t just look like P(text|internet) True, but the same can be said of many things; e.g. biology just looks like P(reproduction|environment), the economy just looks like P(profit|markets), etc. There can still be rich structure inside, and useful abstractions to describe them.

I don't think anyone looks a biological systems and thinks they are well described by P(reproduction|environment). ChatGPT, on the other hand, is P(text|internet) more or less by definition (it's more than that, but not much more).

The fact that we observe what looks like emergent structures such as "understanding", "knowledge" or "reasoning" is fantastic, but it is not in any way incompatible with a P(text|internet) model simply "mimicking" humans.

I agree that exploring the inner workings of these emergent features is interesting in it's own right, but all that glitters is not gold.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#162

Post by antonvs »

belter wrote: post #50
lmm wrote: post #29

Earlier quoted context omitted.

How is that impossible? Plenty of libraries are available for COBOL, especially if you use COBOL.NET

What is the COBOL SDK for AWS?

There are http client libraries for COBOL, and it’s easy to use http to make S3 api calls.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#163

Post by ibejoeb »

on how being trained on code may have given rise to chain-of-thought in natural language:

"...think about how procedure-oriented programming is similar to solving tasks step by step, and how object-oriented programming is similar to decomposing complex tasks into simpler ones."

I get the point, but damn that's not at all how I'd describe those paradigms.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#165

Post by krisoft »

vikp wrote: post #31
srajabi wrote: post #2

Amazing insight, particularly section 6. "- The two important but different abilities of GPT-3.5 are *knowledge* and *reasoning*. Generally, it would be ideal if we could *offload the knowledge part to the outside retrieval system and let the language model only focus on reasoning.* This is because: - The model’s internal knowledge is always cut off at a certain time. The model always needs up-to-date knowledge to an…

It's unclear to me how you could separate knowledge and reasoning: - Reasoning typically requires base knowledge to work from. A side effect of training reasoning is embedding knowledge into the model parameters. - Even if you offload the search portion (either through outputting special tokens that are postprocessed, or applying the model in multiple steps with postprocessing), you still need embedded knowledge for…

I don’t know why you would want to separate them completely. You have some knowledge in your head, store some other in notes on your desk and some in a library (the kind with books in it).

You can make the model write to a local wiki when it encounters new information and read from its own wiki when it feels it needs to store stuff. You can also make it spend time randomly browsing the knowledge base reconsolidating it, reorganising and labeling it.

The architecture of the wiki doesn’t have to be “clever” in any way. It is just an old fasioned database with query function the model can write to and query from.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#166

Post by DebtDeflation »

benreesman wrote: post #147

I have yet to see an output from a big language model that doesn’t just look like P(text|internet). I understand that it’s very easy to ascribe all kinds of qualities to these things, but when the corpus is the Internet, the log likelihood of it sounding like a person is not so different from the corpus sounding like a person. These things are impressive enough without any magical thinking.

The more I play with it, the more it feels like a Google Knowledge Graph result that's been run through a generative model to rewrite the text in a different style.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#167

Post by mensetmanusman »

xeromal wrote: post #120
pixl97 wrote: post #100

Earlier quoted context omitted.

If it's based on frequency of training data, the ex-president will have far more hits in the training corpus. Now, on less talked about topics it doesn't sound any different than what happens with people Q "How old is Tim?" A "Which Tim are you talking about, you didn't give me crap to work with?"

The answer is obviously 62 because that's how old Tim Apple is.

Ha, it might give an age of 7 years or so when the moniker was invented.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#168

Post by astrange »

matheusmoreira wrote: post #101
geraldwhen wrote: post #55

Earlier quoted context omitted.

Competitors will remove the political correctness filters and become more prominent. A Chinese stable diffusion art site is already gaining traction, no filters.

Hope one day we have the ability to run these things on our own computers with 0 restrictions.

Already do. (Is the Chinese site actually based on Stable Diffusion though? It looks pretty different.)

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#169

Post by bheadmaster »

dunefox wrote: post #137
lovecg wrote: post #130

Earlier quoted context omitted.

Fascinating - makes me think of the distinction between dreaming and being awake. Only in the latter state one can tell the difference.

Not quite true, see lucid dreaming.

To be fair to the GP comment, I've personally never experienced lucid dreaming, so for me the distinction holds - when I'm dreaming, I never know I'm dreaming, and mostly I don't even remember my life awake, I only remember whatever's in the context of the dream.

Re: How does GPT obtain its ability? Tracing emergent abilities of language models

#170

Post by kewp »

benreesman wrote: post #147

I have yet to see an output from a big language model that doesn’t just look like P(text|internet). I understand that it’s very easy to ascribe all kinds of qualities to these things, but when the corpus is the Internet, the log likelihood of it sounding like a person is not so different from the corpus sounding like a person. These things are impressive enough without any magical thinking.

Isn't it possible that intelligence is P(words|every sequence of words you've ever heard)?
Post reply on HN