upvote

points

by DrewADesign10 hours ago |
comments
upvoteby K0balt8 hours ago|
next
[-]
It is text prediction. But to predict text, other things follow that need to be calculated. If you can step back just a minute, i can provide a very simple but adjacent idea that might help to intuit the complexity of “ text prediction “ .

I have a list of numbers, 0 to9, and the + , = operators. I will train my model on this dataset, except the model won’t get the list, they will get a bunch of addition problems. A lot. But every addition problem possible inside that space will not be represented, not by a long shot, and neither will every number. but still, the model will be able to solve any math problem you can form with those symbols.

It’s just predicting symbols, but to do so it had to internalize the concepts.

reply
upvoteby qsera5 hours ago|
parent|
[-]
>internalize the concepts.

This gives the impression that it is doing something more than pattern matching. I think this kind of communication where some human attribute is used to name some concept in the LLM domain is causing a lot of damage, and ends up inadvertently blowing up the hype for the AI marketing...

reply
upvoteby cyanydeez10 hours ago|
prev|
next
[-]
There was a paper recently that demonstrated that you can input different human languages and the middle layers of the model end up operating on the same probabilistic vectors. It's just the encoding/decoding layers that appear to do the language management.

So the conclusion was that these middle layers have their own language and it's converting the text into this language and this decoding it. It explains why sometime the models switch to chinese when they have a lot of chinese language inputs, etc.

reply
upvoteby DrewADesign10 hours ago|
parent|
next
[-]
Ok — that sounds more like a theory rather than an open-and-shut causal explanation, but I’ll read the paper.
reply
upvoteby trenchgun7 hours ago|
parent|
[-]
You’re a literature cycle behind. ‘Middle-layer shared representations exist’ is the observed phenomenon; ‘why exactly they form’ is the theory.

You are also confusing ‘mechanistic explanation still incomplete’ with ‘empirical phenomenon unestablished.’ Those are not the same thing.

PS. Em dash? So you are some LLM bot trying to bait mine HN for reasoning traces? :D

reply
upvoteby DrewADesign4 hours ago|
parent|
[-]
Oh, Jesus Christ. I learned to write at a college with a strict style guide that taught us how to use different types of punctuation to juxtapose two ideas in one sentence. In fact, they did/do a bunch of LLM work so if anyone ever used student data to train models, I’m probably part of the reason they do that.

You sound like you’re trying to sound impressive. Like I said, I’ll read the paper.

reply
upvoteby cyanydeez4 hours ago|
parent|
[-]
Congrats on reading.
reply
upvoteby DrewADesign54 minutes ago|
parent|
[-]
Sick burn
reply
upvoteby skydhash8 hours ago|
parent|
prev|
[-]
Pretty obvious when you think that neural networks operate with numbers and very complex formulas (by combining several simple formulas with various weights). You can map a lot of things to number (words, colors, music notes,…) but that does not means the NN is going to provide useful results.
reply
upvoteby DrewADesign4 hours ago|
parent|
[-]
Everything is obvious if you ignore enough of the details/problem space. I’ll read the paper rather than rely on my own thought experiments and assumptions.
reply
upvoteby pennaMan10 hours ago|
prev|
[-]
>It’s literally text prediction, isn’t it?

you are discovering that the favorite luddite argument is bullshit

reply
upvoteby DrewADesign10 hours ago|
parent|
prev|
[-]
Feel free to elucidate if you want to add anything to this thread other than vibes.
reply
upvoteby electroglyph10 hours ago|
parent|
[-]
after you go from from millions of params to billions+ models start to get weird (depending on training) just look at any number of interpretability research papers. Anthropic has some good ones.
reply
upvoteby HumanOstrich9 hours ago|
parent|
next
[-]
> things start to get weird

> just look at research papers

You didn't add anything other than vibes either.

reply
upvoteby Barbing6 hours ago|
parent|
prev|
next
[-]
Interesting, what kind of weird?
reply
upvoteby DrewADesign9 hours ago|
parent|
prev|
[-]
Getting weird doesn’t mean calling it text prediction is actually ‘bullshit’? Text prediction isn’t pejorative…
reply