But what is a GPT? Visual intro to transformers | Chapter 5, Deep Learning
Vložit
- čas přidán 7. 05. 2024
- Unpacking how large language models work under the hood
Early view of the next chapter for patrons: 3b1b.co/early-attention
Special thanks to these supporters: 3b1b.co/lessons/gpt#thanks
To contribute edits to the subtitles, visit translate.3blue1brown.com/
Other recommended resources on the topic.
Richard Turner's introduction is one of the best starting places:
arxiv.org/pdf/2304.10557.pdf
Coding a GPT with Andrej Karpathy
• Let's build GPT: from ...
Introduction to self-attention by John Hewitt
web.stanford.edu/class/cs224n...
History of language models by Brit Cruise:
• ChatGPT: 30 Year Histo...
Paper about examples like the “woman - man” one presented here:
arxiv.org/pdf/1301.3781.pdf
------------------
Timestamps
0:00 - Predict, sample, repeat
3:03 - Inside a transformer
6:36 - Chapter layout
7:20 - The premise of Deep Learning
12:27 - Word embeddings
18:25 - Embeddings beyond words
20:22 - Unembedding
22:22 - Softmax with temperature
26:03 - Up next
------------------
These animations are largely made using a custom Python library, manim. See the FAQ comments here:
3b1b.co/faq#manim
github.com/3b1b/manim
github.com/ManimCommunity/manim/
All code for specific videos is visible here:
github.com/3b1b/videos/
The music is by Vincent Rubinetti.
www.vincentrubinetti.com
vincerubinetti.bandcamp.com/a...
open.spotify.com/album/1dVyjw...
------------------
3blue1brown is a channel about animating math, in all senses of the word animate. If you're reading the bottom of a video description, I'm guessing you're more interested than the average viewer in lessons here. It would mean a lot to me if you chose to stay up to date on new ones, either by subscribing here on CZcams or otherwise following on whichever platform below you check most regularly.
Mailing list: 3blue1brown.substack.com
Twitter: / 3blue1brown
Instagram: / 3blue1brown
Reddit: / 3blue1brown
Facebook: / 3blue1brown
Patreon: / 3blue1brown
Website: www.3blue1brown.com
Edit: The finalized version of the next chapter is out czcams.com/video/eMlx5fFNoYc/video.html
Early feedback on video drafts is always very important to me. Channel supporters always get a view of new videos before their release to help inform final revisions. Join at 3b1b.co/support if you’d like to be part of that early viewing group.
@3Blue1Brown thanks explaining these things - it is very hard for web programmer to undestand math
Grant, this is so good! I've worked in ML about 8 years and this is one of the best descriptions I've seen. Very nicely done. Big 👍👍👍
Is this worth understanding?
@@JohnSegrave sir, could you recomend video analysis framework any video description model?
Could you share the name of the model that is used for text-to-speech generation ? Me and my teammate are working on a Song Translator as a senior design project. This might be very helpful. Thanks in advance :)
Grant casually uploading the best video on Transformers on CZcams
i was expecting froggin electromagnets to be honest :-)
This video was insanely good!
Even having a basic understanding of what it is, this was still extremely helpful!
yeah but it did not talk about transformers in this chapter
I wish I could retweet this post.
I graduated from Computer Science in 2017. Back then, the cutting edge of ML were Recurrent Neural Networks, in which I based my thesis. This video (and I'm sure the rest of this series) just allowed me to catch up to years of advancements in so little time.
I cannot describe how important your teaching style is to the world. I've been reading articles, blogs, papers on embeddings and these topics for years now and I never got it quite like I got it today. In less than 30 minutes.
Imagine a world in which every teacher taught like you. We would save millions and millions of man hours every hour.
You truly have something special with this channel and I can only wish more people started imitating you with the same level of quality and care. If only this became the standard. You'd deserve a Noble Prize for propelling the next thoustand Nobel Prizes.
Second that!
Same, I did a thesis about vectorize word back in 2017 and no one ever talked about the whole vector of word gives rise to meaning and context when you generate phrases.
Too bad since noone was interested in ML back then, I leaned on web development and drop the ML :(
Funny enough, the other 6 videos in this series all came out in 2017, so you probably didn't miss much.
Well transformers were first developed in 2017 so it was the cutting edge exactly when you graduated ^^
This is explained in layman terms, but in reality is more complicated than this
The fact that meaning behind tokens is embedded into this 12000 dimensional space, and you get relationships in terms of coordinates and direction, that exists across topics is mind blowing. Like, Japan -> sushi is similar to Germany -> bratwurst is just so darn neat
I actually went berserk when I saw that part of the video, so friggin cool.
And it makes the absurdly ham fisted model tampering behind debacles like the Gemini launch look even more absurd. I can hear the troglodytes mobbing in the nth dimension.
I‘ve come to this realization long time ago then I want to find isomorphic structures of concepts across different disciplines
@@dayelu2679🤓
You can actually try this out in your nearest large language model, like ChatGPT, CoPilot, Gemini, or Mistral. Just ask it to do vector math on the words. Since there isn't a predefined vector word calculus in English, the LLM defaults to just using a version of its own internal representation, and so it can eke out pretty results. I was able to duplicate Hitler - Germany + Italy = Mussolini and sushi - Japan + Germany = sausage (or bratwurst, bother score highly) in GPT-3.5-Turbo Complete.
It also figured out sushi - Japan + Lebanon = shawarma; sushi - Japan + Korea = kimchi; Hitler - Germany + Spain = Franco; and Hitler - Germany + Russia = Stalin.
This video is gonna blow up. The visualizations will help many people that aren't familiar with NN's or Deep Learning to at least grasp a little bit what is happening under the hood. And with the crazy popularity of LLM's nowadays, this will for sure interest a lot of people
as someone who gave a lot of fellow students lessons in stem field classes i can tell you that the sheer amount of numbers arranged in matrices will immediately shut down the average persons brain...
No, normal people just want to party.
I was trying to understand chatGPT through videos and texts on the Internet. I always said: I wish 3b1b releases a video about it, it's the only way for someone inexperienced to understand, and here it is. Thank you very much for your contributions to youtube!!
no even the other videos are kinda meh, even if youre not inexperienced because they dont go in depth, i feel here people get a nice understanding of the concepts captured by the models instead of just the architecture of the models
It's kind of true, but if I had to recommend a good place to actually understand transformers and even other machine learning things I would definitely recommend StatQuest, its levels of clearly explaining what's going on are very high. But I'm also very excited to see how 3B1B is going to render all that visually as always
I was also just looking into transformers due to their extreme takeover in computer vision!
shoulda just asked chatgpt
NLP specialization by Andrew covered everything 😅
I don't even know how many times I'm going to rewatch this.
True
3B1B doens't need to be saved in watch later folder because all his videos are worth watching later.
What will you set your weights n biases too?
same
This is heaven for visual learners. Animations are correlated smoothly with the intended learning point ...
There's no such thing as visual learners. Other than the blind, all humans are visual creatures. It's heaven for anyone who wants to learn.
@@gorgolyt You are right. The human get input from 5 senses , but 90 percent of the brain receptors are directly connected to optical and auditory nerves. That is where the visual dominates the other senses ... For blind people the auditory dominates...
This guy taught me how to build a neural network from scratch, I was waiting for this video, I even posted a request for it in the subreddit for this channel. I’m very glad this finally exists
The return of the legend! This series is continuing, that is the best surprise of CZcams, thanks Grant, you have no idea how much the young population of academia is indebted to you.
I'm 26, young engineer. Thinking the same. Well said.
I liked your comment because I'm sure you're right but don't be ageist! 😊 Us olds love him too!
Young and seniors alike
I have been working on transformers for the past few years and this is the greatest visualization of the underlying computation that I have seen. Your videos never disappoint!!
So if we "stop" you... then we avoid judgement day? We should meet for coffee
@@brian8507 "judgement day" 😭
@@brian8507 bro's got underlying psychological issues
I agree with you . Visualization is perfect way to understanding transformer architecture. Specifically attention mechanism
@@giacomobarattini1130its later than you thinj
I wish i had a friend as passionate as this channel is. It's like finding my family I've always wanted to have
I wish the same thing
become friends already you both
here we are 3b1bro now
Lets become friends
I have always been blown away by how great your visualization is for explaining ML concepts. Thanks a lot!
You *must* turn the linguistic vector math bit into a short. -Japan+sushi+germany=bratwurst is pure gold.
I am slightly offended it did not result in "Fischbrötchen".
@@XMysticHerox It was trained in English words only.
@@marshmellominiapple ChatGPT supports 95 languages. Not all equally well. But as a German yes it works just as well with german as it does with english.
@@marshmellominiapple Word2Vec and other vector embeddings of words like glove or whatever don't care about language. They don't "understand" the meaning of the words, they just eventually find patterns in unstructured data to create the embeddings. It works in any language and GPT has a ton of other languages in its training data
You can actually try this out in your nearest large language model, like ChatGPT, CoPilot, Gemini, or Mistral. Just ask it to do vector math on the words. Since there isn't a predefined vector word calculus in English, the LLM defaults to just using a version of its own internal representation, and so it can eke out pretty results. I was able to duplicate Hitler - Germany + Italy = Mussolini and sushi - Japan + Germany = sausage (or bratwurst, bother score highly) in GPT-3.5-Turbo Complete.
It also figured out sushi - Japan + Lebanon = shawarma; sushi - Japan + Korea = kimchi; Hitler - Germany + Spain = Franco; and Hitler - Germany + Russia = Stalin.
Writing my first academically published paper on AI rn and I have to say as a engineer in this space, this is one of the most complete and well nuanced explanations of these tools. Gold, nay platinum standard for educational content on this topic for decades to come.
Yes. I really hope that he gets some lifetime achievement massive-footprint-in-a-good-sense type of award in the MINT Edu field.
2 years ago I started studying transformers, backpropagation and the attention mechanism. Your videos were a corner stone for my understanding of those concepts!
And now, partially thanks to you, I can say: “yeah, relatively smooth to understand”
Thank you! You're so late 3Blue1Brown, it took me 10 hours of videos + blogs last year to understand what a transformer is! This is the long waited video! I'm sending this to all my friends.
Grant shows just how creative you can get with linear algebra. Who would have guessed language (?!) was within its reach?
Look up "Word2Vec", it's an interestingly explored idea.
Linear algebra would not be enough, but a nonlinear activation function (even one as simple as max(x, 0)) makes it enough to approximate anything you want just by adding more neurons!
Given words are descriptors and numbers are just arbitrarily precise adjectives... aka descriptions...
@@Jesin00 Yes, lin alg alone isn't enough.
Well ML uses linear algebra and he just explains it
Its astonishing, amazing that this kind of info and explaination quality is available for free, this is way better than a University would explain it
Universities are buildings. Buildings can't talk. Therefore, they cannot explain.
Man! You never fail to enlighten, entertain, and inspire us, nor do we get enough of your high-quality, yet very digestible, content! Thank you, Grant!
It's absolutely ridiculous how many aspects of this topic finally clicked for me in this intro video already. This was incredibly well explained an I'm so thrilled for the next chapters. Thank you very much, Grant!
Straight up the best video on this topic. The idea that the dimensions of the embedding space represent different properties of a token that can be applied across tokens is just SO cool!
i felt that
orienting and ordering the space (called the 'latent' space) so that the most significant directions come first is called 'principal component analysis' (useful for giving humans the reigns to some degree since we get to turn those knobs and see something interesting but vaguely predictable happen)
I agree. I starting writing about that in a comment about 2 seconds into the video before I knew how well he was going to cover it since it's usually glossed over way too much in other introductions to these topics.
I always thought when people in the media say, "NO ONE actually understands how chat GPT works" they were lying, but no one was ever able to explain it in layman's terms regardless. I feel like this video is exactly the kind of digestible info that people need, well done.
Machine learning engineer here - plenty of people understand how the architecture of chatGPT works on a high level. When people in the media say that, what they mean is that nobody understands the underlying processing that the parameters are using to go from a list of tokens to a probability distribution over possible next tokens.
It's not a lie, it's just not very precise. No one can tell you exactly why one model decided the next word is "the" while another decided the next word is "a" and in that sense no one understands how a particular model works. The mechanism for how you train and run the model are understood however.
@@kevinscalesyeah just like any deep ANN
Think of it as the difference between knowing how genetics and DNA and replication works vs. knowing why a specific nucleotide in the human genome is adenine rather than guanine.
There is an entire field of machine learning research dedicated to understanding how neural nets work beyond the architecture called AI interpretability.
No one fully understands what the learned parameters mean. Many people understand the process by which they were learned.
You are such an AMAZING teacher. I feel like you've really given thought to the learners perception and are kind enough to take the time and address asides and gotchas while you meticulously build components and piece them together all with a very natural progression that's moving towards "something" (hopefully comprehension). Thank you so much for your time, effort, and the quality of your work.
I listened to my professor explaining the crazy equation of softmax for a semester already, and you explained it so well with how temperature also plays a role there. Big RESPECT!
18:45 This the the clearest layman explanation of how attention works that I've ever seen. Amazing.
here's to hoping this is not an April fools
it is - you would be a fool to not watch this video
It's 2nd April here
@@anuragpranavEven if you already know the subject? 😂
@@TheUnderscore_ it's never a bad idea to review what you know
@@TheUnderscore_ you are almost certainly limiting what you might know with that approach
I've been following this high-quality channel for years. And I don't know how it continues to improve over time.
Thank you for your hard work of popularization of complex notions and your work of aid to intuition with increible visual representations,
really I take my hat off to you, once more : Thank You
Ooh attention ! Looking forward to it. Thank for feeding our attention span with such quality visualisations. From gradients to signifiant coordinates in concept space your visual language keeps getting more refined and synthetic. The adventure through back propagation was rather mind-bending. This series is amazing.
This series should continue. I thought it was dead after the 4th video. Lots of love and appreciation for your work
Im a mechanical engineering student, but I code machine learning models for fun. I was telling my girlfriend just last night that your series on dense neural networks is the best to gain an intuitive understanding on the basic architecture of neural networks. You have no idea what a pleasant surprise it was to wake up to this!
good man
It doesn't have to be just for fun. I was also in Mechanical Engineering, picked a master in control theory. And now I get to use Deep learning and NN's for intelligent control systems. Where you learn a model or a controller by making use of machine learning
The quality of these videos and depth of openings of the deeper meaning is simply mind blowing
Your teaching skills are beyond compare. The effort you put into your videos clearly shows.
I appreciate that you explain tokenization correctly and the usefulness of simplifying it. Many explanations skip all that and just state that the tokens are words.
Apart from the fact that tokens CAN actually be longer than a word, too. :) Sub-word token does not mean that tokens must be smaller than a word.
There is a related idea here in how Roombas navigate houses. They clearly are forming a map of your house in their memory, but there is no guarantee they see it the same way we do i.e. the different zones they see in your house may not correspond nicely to the actual rooms in the house. In the end, though, it doesn't really matter, as long as the job gets done correctly
We need more Computer Science education like this! Amazing 🔥
Honestly I hope that in future, AI can produce such great content. This will probably tend to take a couple of years more, but I guess its possible. Even better: You got your own Curriculum based on your strengthens and weaknesses. For me this would be a combination of fireship and 3blue1brown content...
This is absolutely one of the best videos for explaining the workings of LLMs. Love the visualisation and the innate ease with which the concepts were explained.
Hats off!!
This channel is so good!!
The way such complicated topics are broken down and explained is really of the highest standard.
Please never stop making videos!
You made my day!! This topic was taught at my grad school and I needed some intuition today and you have uploaded the video!!! It seems you heard me!!Thanks a ton!! Please upload video of Vision Transformers, if possible
The idea of describing a vector basis as a long list of questions you need to answer is exactly the teaching tool I needed in my kit!! I love that perspective!
Thank you so much for this video. I have watched many on the topic and they teach you some parts, but you connected all the missing dots in a very effective way. Thanks.
Brilliant! Although I already knew pretty much everything presented here, it brings a great visual touch to all the concepts which helps sink those even deeper into the mind. Terrific job as usual! Looking forward to the next chapters of the series!
this channel is so good at making math interesting, all my friends think im wack for enjoying math videos but its not hard to enjoy when you make them like this
The best video explaining the transformer architecture that I've seen so far... and there are really good videos covering this topic. Thank you!!
You never cease to amaze me. This is a must watch for any engineer or data scientist. You deserve to be the top one youtube channel. Thank you brother
This is genuinely one of the best pieces of science / tech communication I have ever seen. Well done and thank you!
This couldn't have come at a better time for me! I'm very excited for this continuation of the series. Thanks Grant!
Im in highschool, and i only knew broken pieces of how it works , but you really connected all the pieces together and added the missing ones
Your videos are the only ones I watch multiple times to get the hang of it. But the way you visualize and explain it makes it way more enjoyable and interesting. Love your videos ❤
Thank you so much for the immense work and talent that goes into your videos!
I can't believe Sam Altman has become a billionaire just by multiplying some matrices
You'd be surprised at how many billionaires got there from multiplying some matrices
Also, gets away with calling it "machine learning".
It's too much reduction, he added value on a higher level. But yeah, when you look deep enough, everything stops looking like magic.
@@tiborsaas And that is a good thing in many cases, it casts away illogical fears when you understand that there is no any kind of magic or thinking behind this. In practice it is just overhpyed guessing machine what word normally might come after X.
@@user-gw3yb3ki6w this concept comes from 2017. We should actually be very very worried and keeping our eye closely on the progress that AI is making. The amount of progress they have made since the 2017 paper 📝 “Attention is all you need “ is insane.
I usually gloss over your videos as it gets more technical. But whatever you did this time, kept me hooked the whole time.
This is by far the most organized explanation i've seen about transformers.
I know the material of this chapter very well. Still, I watched it in its entirety just for the pleasure of watching a masterful presentation, the restful and authoritative cadence of the voice, and the gorgeous animation. Well done, Grant, yet again.
So you’re telling me computer models can quantify “a few” or “some” based on how close the value is to a given word of a number from its usage from training data?
I love this
Well, a bit.
Well it can encode any semantic meaning only really limited by the number of parameters and quality of training data.
@@XMysticHerox quantity*
Gold resource! Great communication skill! Always a pleasure to listen to. Thank you
The visual of this video are next level, really help the comprehension of the subject
Amazing work, your simple explanations in other videos in this series really helped me get a better understanding of what my masters classes were covering. Glad to see you’re continuing this series! ❤
Your explanation tends to always be the best! Love how you visualise all these.
SANTIAGO DE CHILE MENTIONED!!! 🗣️🔥🔥🔥 WHAT THE HELL IS CLEAN AIR!!?!?!?!? 🗣️🗣️🗣️🔥🔥
Wow!! I've no prep in math, analisis, programmig, machie learning, etc... and can grab the concept!
What a MIRACLE have you done here! Good work! 🤯
ok, you read my mind. From all of the channels, I am so glad, it's you explaining Transformers.
I wish I had you as a teacher. You make math so much more fun than I know it already❤
It amazing that this knowledge is free, really learned a lot from this short session. Definitely will binge watch your videos.
Seriously amazing content. The way you present things visually is excellently done.
this is the best series from you by far. excited for its revival
Deep learning is back baby!
A short 6 year 5 month wait!
Love that bit about the word embeddings and how each direction in that high dimensional space carries some semantic meaning to a certain degree. Haven't heard something i found that interesting in years!
Wow I think this is the best YT video I have seen this year. I understood most of it and I think now I am going to really dive into this topic. Thanks 3B1B
I love 3Blue1Brown series ..the linear algebra series was really the state-of-the-art and recognized globally for AI enthusiasts like myself. Now hot topics such as Transformers and GenAI , this is really the best explanation by far. Its short and precise and that's what we want.
Most exciting part of my week by far
I’m a professional in this field, but always there is so much more to learn and understand. This is so incredibly high level and detailed at the same time, with so much intuition. Masterfully presented, thank you ❤
Thank you so much. We were just learning about transformer architecture in our deep learning course. This was really helpful for visualizing word embeddings!
OH MY GOODNESS
Your timing is just right! I'm learning about deep neural nets and transformers will be my next topic this week.
I'M SO EXCITED, I JUST CAN'T HIDE IT!
I'M ABOUT TO LOSE MY MIND AND I THINK I LIKE IT!
Without watching i can say that this is going to be the best transformers video on yt
Right you are.
It is
This pure gold. Amazing. The quality of your videos is the highest on YT.
Hi Grant, when I first saw your original 5 deep learning videos I was in my second to last year of high-school, they were my first introduction to ML and deep learning, and they played a part in me choosing to study Computer Science as an undergraduate. Now, five years later I am working on my masters thesis where I am using Vision Transformers :)
The BEST AI video. Highly recommended !
Hello 3b1b, I wanted to say a huge thank you for this specific video. This was exactly what I've been needing. Every now and again, I thought to myself, as someone who's been interested in machine learning for my whole adult life, that I should really get a deep understanding of how a transformer works, to the point that I could implement a functional, albeit not efficient, one myself.
Well, I'm on my way to that, this is at least a great introduction (and knowing your channel I really mean GREAT), and I really wanted to thank you for that!
I know this is not much, but I'm not in a position to support this channel in a more meaningful way at the moment.
Anyways, take care, and thanks again!
I'm glad you enjoyed. In case some how you haven't already come across them, I'd recommend the videos Andrej Karpathy does on coding up a GPT. In general, anything he makes is gold.
6:18 - Remarkably clear explanation expanded upon the very, very nice graphics. Another 3B1B success story! I finally have my head wrapped around GPT...
The word embedding difference example is.. incredible
I never thought about it this way
Thank you so much for this!
Hang on, I thought this series was done! I'm delighted!
Damit nice April fools joke, I got fooled into learning something.
Just brillant.. there is no video in CZcams that explains concepts in such a clever way
thank you so much. as a highschool student who's deeply intrigued by LLMs and deep learning, this was so much better than me trying to interpret the "attention is all you need" paper myself (with LLMs to help, ironically) haha.
this is hands down the best resource on the transformer architecture and deep learning I've ever found - and I've been through a LOT.
thank you :)
YOU DID IT!
I emailed you about this video idea about 8 months ago and I've been patiently waiting for you to release this since!
wow, great idea!
YOU DID IT JACKSON! I texted you to email him this idea about 9 months ago. Now the bab- video is there!
CHILE MENTIONED 🇨🇱🇨🇱❤️❤️🇨🇱🇨🇱🇨🇱 COME TO SANTIAGO GRANT!!!
the 'embeddings beyond words' segment levelled up my understanding of how machine learning 'thinks'. thank you!
Thank you for your contributions. This channel is an absolute goldmine for people like me!
As an ML researcher this is an amazing video ❤. But please allow me to nitpick a little at 21:45
It’s important to note that while the “un-embedding layer” of a Transformer typically have a different set of weights from the embedding layer, in OpenAI’s GPT model each vector for each word in the un-embedding layer is exactly the same vector as ones in the embedding layer.
This is not the case for Transformer models that has the output be in a different domain than the input (e.g, translating to a different language), but since the video is specifically talking about GPT. This is the specific of the implementation detailed in the “Improving Language Understanding by Generative Pre-Training” paper by OpenAI.
The reusing weights make sense here because each the vector from the embedding is a sort of “context free” representation of the word. So there is not need to learn another set of weights.
I started my deep learning journey from your original videos on deep learning. They inspired me to work in this field. I am about to start my first internship as a researcher in this field. Thank you 3blue1brown for this.
Also this is the best video that I have seen through my many hundred videos from when I was stuck in tutorial hell on many of these concepts
Just in time to be replaced by them >:).
And again and again and again in the pool of Million explenations, Yours is the one that makes the topic understandable for everyone who is able to watch and listen. Great work!
This is incredible! Thank you so much for making this so simple to understand. Amazing amazing work !
I'm a simple man. I see 3B1B has released a new video, I click...
Before I clicked on this video, I thought a Transformer was a fictional machine-like species that was being used as toys. Now I know that transformers are much more different. Thanks Grant.
They’re more than meets the eye. ;)
The most interesting video I've watched on CZcams EVER! Very well done explaining this complex topic.
Great video! The explanation with the animation makes it easy to understand such a complex topic
3blue1brown released a normal video today. So did Numberphile. So did nearly all the channels in my subsd. There's no wacky bullshit on the google homepage. No stupid gimmick feature in Maps. Have we done it? Have we finally killed off the lamest holiday? Is it finally dead?
The different sampling has to do with the search algorithm, like beam search, or any search involving topk or some tally of probabilities for the final score of the output. Any temperature will not change that the most probable token is the most probable token, so in a greedy search the temperature does not affect the output. This is a very common misconception, I'm a bit disappointed that it was slightly misleading here.
agree, what he said wasn’t logically complete and didn’t really make sense because of it
To be clear, rest of the video was great!
Excellent breakdown and amazing visualisations. Thanks for this huge effort.
I've no words to express my gratitude for you, for making these. Thanks! You're awesome!