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

The universal weight subspace hypothesis

arxiv.org

Post reply on HN
Print view View original

51–60 of 146 posts

Re: The universal weight subspace hypothesis

#51

Post by scotty79 »

altairprime wrote: post #19

For those trying to understand the most important parts of the paper, here's what I think is the most significant two statements, subquoted out of two (consecutive) paragraphs midway through the paper: > we selected five additional, previously unseen pretrained ViT models for which we had access to evaluation data. These models, considered out-of-domain relative to the initial set, had all their weights reconstructed…

16 seems like a suspiciously round number ... why not 17 or 13? ... is this just result of some bug in the code they used to do their science?

or is it just that 16 was arbitrarily chosen by them as close enough to the actual minimal number of dimensions necessary?

Re: The universal weight subspace hypothesis

#52

Post by N_Lens »

altairprime wrote: post #19

For those trying to understand the most important parts of the paper, here's what I think is the most significant two statements, subquoted out of two (consecutive) paragraphs midway through the paper: > we selected five additional, previously unseen pretrained ViT models for which we had access to evaluation data. These models, considered out-of-domain relative to the initial set, had all their weights reconstructed…

If models naturally occupy shared spectral subspaces, this could dramatically reduce

- Training costs: We might discover these universal subspaces without training thousands of models

- Storage requirements: Models could share common subspace representations

Re: The universal weight subspace hypothesis

#53

Post by brillcleaner »

MarkusQ wrote: post #8
canjobear wrote: post #3

What's the relationship with the Platonic Representation Hypothesis?

From what I can tell, they are very closely related (i.e. the shared representational structures would likely make good candidates for Platonic representations, or rather, representations of Platonic categories). In any case, it seems like there should be some sort of interesting mapping between the two.

My first thought was that this was somehow distilling universal knowledge. Platonic ideals. Truth. Beauty. Then I realized- this was basically just saying that given some “common sense”, the learning essence of a model is the most important piece, and a lot of learned data is garbage and doesn’t help with many tasks. That’s not some ultimate truth, that’s just optimization. It’s still a faulty LLM, just more efficient for some tasks.

Re: The universal weight subspace hypothesis

#54

Post by acjohnson55 »

EvanAnderson wrote: post #12
VikingCoder wrote: post #5

I find myself wanting genetic algorithms to be applied to try to develop and improve these structures... But I always want Genetic Algorithms to show up in any discussion about neural networks...

I have a real soft spot for the genetic algorithm as a result of reading Levy's "Artificial Life" when I was a kid. The analogy to biological life is more approachable to my poor math education than neural networks. I can grok crossover and mutation pretty easily. Backpropagation is too much for my little brain to handle.

You can definitely understand backpropagation, you just gotta find the right explainer.

On a basic level, it's kind of like if you had a calculation for aiming a cannon, and someone was giving you targets to shoot at 1 by 1, and each time you miss the target, they tell you how much you missed by and what direction. You could tweak your calculation each time, and it should get more accurate if you do it right.

Backpropagation is based on a mathematical solution for how exactly you make those tweaks, taking advantage of some calculus. If you're comfortable with calculus you can probs understand it. If not, you might have some background knowledge to pick up first.

Re: The universal weight subspace hypothesis

#55

Post by hn_throwaway_99 »

I read the abstract (not the whole paper) and the great summarizing comments here.

Beyond the practical implications of this (i.e. reduced training and inference costs), I'm curious if this has any consequences for "philosophy of the mind"-type of stuff. That is, does this sentence from the abstract, "we identify universal subspaces capturing majority variance in just a few principal directions", imply that all of these various models, across vastly different domains, share a large set of common "plumbing", if you will? Am I understanding that correctly? It just sounds like it could have huge relevance to how various "thinking" (and I know, I know, those scare quotes are doing a lot of work) systems compose their knowledge.

Re: The universal weight subspace hypothesis

#56

Post by inciampati »

The authors study a bunch of wild low rank fine tunes and discover that they share a common... low rank! ... substructure which is itself base model dependent. Humans are (genetically) the same. You need only a handful of PCs to represent the cast majority of variation. But that's because of our shared ancestry. And maybe the same thing is going on here.

Re: The universal weight subspace hypothesis

#57

Post by gedy »

hn_throwaway_99 wrote: post #55

I read the abstract (not the whole paper) and the great summarizing comments here. Beyond the practical implications of this (i.e. reduced training and inference costs), I'm curious if this has any consequences for "philosophy of the mind"-type of stuff. That is, does this sentence from the abstract, "we identify universal subspaces capturing majority variance in just a few principal directions", imply that all of th…

It could, though maybe "just" in a similar way that human brains are the same basic structure.

Re: The universal weight subspace hypothesis

#58

Post by modeless »

This seems confusingly phrased. When they say things like "500 Vision Transformers", what they mean is 500 finetunes of the same base model, downloaded from the huggingface accounts of anonymous randos. These spaces are only "universal" to a single pretrained base model AFAICT. Is it really that surprising that finetunes would be extremely similar to each other? Especially LoRAs?

I visited one of the models they reference and huggingface says it has malware in it: https://huggingface.co/lucascruz/CheXpert-ViT-U-MultiClass

Re: The universal weight subspace hypothesis

#59

Post by energy123 »

mlpro wrote: post #36
nothrowaways wrote: post #30

> Principal component analysis of 200 GPT2, 500 Vision Transformers, 50 LLaMA- 8B, and 8 Flan-T5 models reveals consistent sharp spectral decay - strong evidence that a small number of weight directions capture dominant variance despite vast differences in training data, objectives, and initialization. Isn't it obvious?

Not really. If the models are trained on different dataset - like one ViT trained on satellite images and another on medical X-rays - one would expect their parameters, which were randomly initialized to be completely different or even orthogonal.

Every vision task needs edge/contrast/color detectors and these should be mostly the same across ViTs, needing only a rotation and scaling in the subspace. Likewise with language tasks and encoding the basic rules of language which are the same regardless of application. So it is no surprise to see intra-modality shared variation.

The surprising thing is inter-modality shared variation. I wouldn't have bet against it but I also wouldn't have guessed it.

I would like to see model interpretability work into whether these subspace vectors can be interpreted as low level or high level abstractions. Are they picking up low level "edge detectors" that are somehow invariant to modality (if so, why?) or are they picking up higher level concepts like distance vs. closeness?

Re: The universal weight subspace hypothesis

#60

Post by themaxice »

hn_throwaway_99 wrote: post #55

I read the abstract (not the whole paper) and the great summarizing comments here. Beyond the practical implications of this (i.e. reduced training and inference costs), I'm curious if this has any consequences for "philosophy of the mind"-type of stuff. That is, does this sentence from the abstract, "we identify universal subspaces capturing majority variance in just a few principal directions", imply that all of th…

Somewhat of a tangent, but if you enjoy the philosophy of AI and mathematics, I highly recommend reading Gödel, Escher, Bach: an Eternal Golden Braid by D. Hofstadter. It is primarily about the Incompleteness Theorem, but does touch on AI and what we understand as being an intelligence
Post reply on HN