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

The Illustrated Transformer

jalammar.github.io

Post reply on HN
Print view View original

21–30 of 92 posts

Re: The Illustrated Transformer

#21

Post by nospice »

Koshkin wrote: post #16

(Going on a tangent.) The number of transformer explanations/tutorials is becoming overwhelming. Reminds me of monads (or maybe calculus). Someone feels a spark of enlightenment at some point (while, often, in fact, remaining deeply confused), and an urge to share their newly acquired (mis)understanding with a wide audience.

So?

There's no rule that the internet is limited to a single explanation. Find the one that clicks for you, ignore the rest. Whenever I'm trying to learn about concepts in mathematics, computer science, physics, or electronics, I often find that the first or the "canonical" explanation is hard for me to parse. I'm thankful for having options 2 through 10.

Re: The Illustrated Transformer

#22

Post by roadside_picnic »

profsummergig wrote: post #2

Haven't watched it yet... ...but, if you have favorite resources on understanding Q & K, please drop them in comments below... (I've watched the Grant Sanderson/3blue1brown videos [including his excellent talk at TNG Big Tech Day '24], but Q & K still escape me). Thank you in advance.

It's just a re-invention of kernel smoothing. Cosma Shalizi has an excellent write up on this [0].

Once you recognize this it's a wonderful re-framing of what a transformer is doing under the hood: you're effectively learning a bunch of sophisticated kernels (though the FF part) and then applying kernel smoothing in different ways through the attention layers. It makes you realize that Transformers are philosophically much closer to things like Gaussian Processes (which are also just a bunch of kernel manipulation).

0. http://bactra.org/notebooks/nn-attention-and-transformers.ht...

Re: The Illustrated Transformer

#23

Post by kadushka »

Koshkin wrote: post #16

(Going on a tangent.) The number of transformer explanations/tutorials is becoming overwhelming. Reminds me of monads (or maybe calculus). Someone feels a spark of enlightenment at some point (while, often, in fact, remaining deeply confused), and an urge to share their newly acquired (mis)understanding with a wide audience.

Maybe so, but this particular blog post was the first and is still the best explanation of how transformers work.

Re: The Illustrated Transformer

#24

Post by roadside_picnic »

red2awn wrote: post #3
profsummergig wrote: post #2

Haven't watched it yet... ...but, if you have favorite resources on understanding Q & K, please drop them in comments below... (I've watched the Grant Sanderson/3blue1brown videos [including his excellent talk at TNG Big Tech Day '24], but Q & K still escape me). Thank you in advance.

Implement transformers yourself (ie in Numpy). You'll never truly understand it by just watching videos.

I personally don't think implementation is as enlightening as far as really understanding what the model is doing as this statement implies. I had done that many times, but it wasn't until reading about the relationship to kernel methods that it really clicked for me what is really happening under the hood.

Don't get me wrong, implementing attention is still great (and necessary), but even with something as simple as linear regression, implementing it doesn't really give you the entire conceptual model. I do think implementation helps to understand the engineering of these models, but it still requires reflection and study to start to understand conceptually why they are working and what they're really doing (I would, of course, argue I'm still learning about linear models in that regard!)

Re: The Illustrated Transformer

#25

Post by roadside_picnic »

krat0sprakhar wrote: post #12
red2awn wrote: post #3

Earlier quoted context omitted.

Implement transformers yourself (ie in Numpy). You'll never truly understand it by just watching videos.

Do you have a tutorial that I can follow?

The most valuable tutorial will be translating from the paper itself. The more hand holding you have in the process, the less you'll be learning conceptually. The pure manipulation of matrices is rather boring and uninformative without some context.

I also think the implementation is more helpful for understanding the engineering work to run these models that getting a deeper mathematical understanding of what the model is doing.

Re: The Illustrated Transformer

#26

Post by amelius »

roadside_picnic wrote: post #17
qoez wrote: post #13

Earlier quoted context omitted.

Don't really see why you'd need to understand how the transformer works to do LLMs at work. LLMs is just a synthetic human performing reasoning with some failure modes that in-depth knowledge of the transformer interals won't help you predict what they are (just have to use experience with the output to get a sense, or other peoples experiments).

In my experience this is a substantial difference in the ability to really get performance in LLM related engineering work from people who really understand how LLMs work vs people who think it's a magic box. If your mental model of an LLM is: > a synthetic human performing reasoning You are severely overestimating the capabilities of these models and not realizing potential areas of failure (even if your prompt work…

But that AI engineer who is implementing speculative decoding is still just doing basic plumbing that has little to do with the actual reasoning. Yes, he/she might make the process faster, but they will know just as little about why/how the reasoning works as when they implemented a naive, slow version of the inference.

Re: The Illustrated Transformer

#27

Post by zkmon »

I think the internal of transformers would become less relevant like internal of compilers, as programmers would only care about how to "use" them instead of how to develop them.

Re: The Illustrated Transformer

#28

Post by machinationu »

roadside_picnic wrote: post #17
qoez wrote: post #13

Earlier quoted context omitted.

Don't really see why you'd need to understand how the transformer works to do LLMs at work. LLMs is just a synthetic human performing reasoning with some failure modes that in-depth knowledge of the transformer interals won't help you predict what they are (just have to use experience with the output to get a sense, or other peoples experiments).

In my experience this is a substantial difference in the ability to really get performance in LLM related engineering work from people who really understand how LLMs work vs people who think it's a magic box. If your mental model of an LLM is: > a synthetic human performing reasoning You are severely overestimating the capabilities of these models and not realizing potential areas of failure (even if your prompt work…

speculative decoding is 1+1

transformer attention is integrals

Re: The Illustrated Transformer

#29

Post by Koshkin »

qoez wrote: post #13
gustavoaca1997 wrote: post #8

I have this book. Really a life savior to help me catching up a few months ago when my team decided to use LLMs in our systems.

Don't really see why you'd need to understand how the transformer works to do LLMs at work. LLMs is just a synthetic human performing reasoning with some failure modes that in-depth knowledge of the transformer interals won't help you predict what they are (just have to use experience with the output to get a sense, or other peoples experiments).

> is just a synthetic human performing reasoning

The future is now! (Not because of "a synthetic human" per se but because of people thinking of them as something unremarkable.)

Re: The Illustrated Transformer

#30

Post by libraryofbabel »

I read this article back when I was learning the basics of transformers; the visualizations were really helpful. Although in retrospect knowing how a transformer works wasn't very useful at all in my day job applying LLMs, except as a sort of deep background for reassurance that I had some idea of how the big black box producing the tokens was put together, and to give me the mathematical basis for things like context size limitations etc.

I would strongly caution anyone who thinks that they will be able to understand or explain LLM behavior better by studying the architecture closely. That is a trap. Big SotA models these days exhibit so much nontrivial emergent phenomena (in part due to the massive application of reinforcement learning techniques) that give them capabilities very few people expected to ever see when this architecture first arrived. Most of us confidently claimed even back in 2023 that, based on LLM architecture and training algorithms, LLMs would never be able to perform well on novel coding or mathematics tasks. We were wrong. That points towards some caution and humility about using network architecture alone to reason about how LLMs work and what they can do. You'd really need to be able to poke at the weights inside a big SotA model to even begin to answer those kinds of questions, but unfortunately that's only really possible if you're a "mechanistic interpretability" researcher at one of the major labs.

Regardless, this is a nice article, and this stuff is worth learning because it's interesting for its own sake! Right now I'm actually spending some vacation time implementing a transformer in PyTorch just to refresh my memory of it all. It's a lot of fun! If anyone else wants to get started with that I would highly recommend Sebastian Raschka's book and youtube videos as way into the subject: https://github.com/rasbt/LLMs-from-scratch .

Has anyone read TFA author Jay Alammar's book (published Oct 2024) and would they recommend it for a more up-to-date picture?

Post reply on HN