E33: The Tiny Model Revolution with Ronen Eldan and Yuanzhi Li of Microsoft Research

Nathan Labenz sits down with Ronen Eldan and Yuanzhi Li of Microsoft Research to discuss the small natural language dataset they created called TinyStories. Tiny Stories is designed to reflect the full richness of natural language while still being small to support research with modest compute budgets. Using this dataset, they began to explore aspects of language model performance, behavior, and mechanism by training a series of models that range in size from just 1 million to a maximum of 33 million parameters – which is still just 2% the scale of GPT-2. In this conversation, Nathan, Ronen, and Yuanzhi touch on LM reasoning, emergence, interpretability, and what understanding can be extended to LLMs. RECOMMENDED PODCAST: The HR industry is at a crossroads. What will it take to construct the next generation of incredible businesses – and where can people leaders have the most business impact? Hosts Nolan Church and Kelli Dragovich have been through it all, the highs and the lows – IPOs, layoffs, executive turnover, board meetings, culture changes, and more. With a lineup of industry vets and experts, Nolan and Kelli break down the nitty-gritty details, trade offs, and dynamics of constructing high performing companies. Through unfiltered conversations that can only happen between seasoned practitioners, Kelli and Nolan dive deep into the kind of leadership-level strategy that often happens behind closed doors. Check out the first episode with the architect of Netflix’s culture deck Patty McCord. https://link.chtbl.com/hrheretics LINKS: Tiny Stories paper: https://huggingface.co/papers/2305.07759 TIMESTAMPS: (00:00) Episode Preview (07:12) The inspiration for the Tiny Stories project (15:07) Sponsor: Omneky (15:44) Creating the Tiny Stories dataset (21:27) GPT-4 vs GPT-3.5 (24:13) Did the TinyStories team try any other versions of GPT-4 (29:23) Curriculum models and weirder curriculums (35:34) What does reasoning mean? (46:27) What does emergence mean? (01:01:44) The curriculum development space (01:11:40) The similarities between models and human development (01:20:12) Fewer layers vs. more layers (01:29:22) Attention heads (01:33:40) Semantic attention head (01:36:54) Neuron technique used in developing the TinyStories model (01:52:20) Interpretability work that inspires Ronen and Yuanzhi TWITTER: @CogRev_Podcast @EldanRonen (Ronen) @labenz (Nathan) @eriktorenberg (Erik) SPONSORS: Shopify is the global commerce platform that helps you sell at every stage of your business. Shopify powers 10% of ALL eCommerce in the US. And Shopify's the global force behind Allbirds, Rothy's, and Brooklinen, and 1,000,000s of other entrepreneurs across 175 countries. From their all-in-one e-commerce platform, to their in-person POS system – wherever and whatever you're selling, Shopify's got you covered. With free Shopify Magic, sell more with less effort by whipping up captivating content that converts – from blog posts to product descriptions using AI. Sign up for $1/month trial period: https://shopify.com/cognitive Thank you Omneky for sponsoring The Cognitive Revolution. Omneky is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work, customized across all platforms, with a click of a button. Omneky combines generative AI and real-time advertising data. Mention "Cog Rev" for 10% off. This show is produced by Turpentine: a network of podcasts, newsletters, and more, covering technology, business, and culture — all from the perspective of industry insiders and experts. We’re launching new shows every week, and we’re looking for industry-leading sponsors — if you think that might be you and your company, email us at [email protected]. Music Credit: MusicLM More show notes and reading material released in our Substack: https://cognitiverevolution.substack.com

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  One of the most important abilities for generative models is to be able to speak coherent English. Her mom didn't let her have a dog, so she asked for a... And when you try to autocomplete this, now, you know, the most common noun that you've seen so far, also the most, the proximate one, is dog, not cat. Dog already appears twice in the sentence. Even GPT-2 XL, which has 1.5 billion parameters, its most likely completion is still dog.

• 0:43Nathan Labenz?

  I'm just gonna read one tiny story, just straight out of the paper, because I think that will help people understand like, what this dataset, you know, ultimately is. Tom has a big pot of soup. He wants to share it with Jane. Jane takes a spoonful of soup, but then she makes a face. The soup is, that's the prompt, and then you show a completion, very bitter. She does not like it. She says, "I don't like this soup, it is too bitter." He looks around the kitchen and finds some bread and cheese. He puts them on the table and says, "Here Jane, you can have some bread and cheese. They are not bitter, they are sweet and yummy." Jane is happy. She says, "Thank you, Tom, you are a good friend. I like bread and cheese, they are not bitter."

• 1:25Nathan Labenz?

  Hello and welcome to The Cognitive Revolution, where we interview visionary researchers, entrepreneurs, and builders working on the frontier of artificial intelligence. Each week, we'll explore their revolutionary ideas, and together, we'll build a picture of how AI technology will transform work, life, and society in the coming years. I'm Nathan Labenz, joined by my co-host, Erik Torenberg. Hello and welcome back to The Cognitive Revolution. Today's episode is great for anyone who really wants to deepen their understanding of and intuition for how language models really work. Certainly, as measured by how much I learned in the course of the conversation, it's one of our very best. Our guests, Ronan Eldan and Yuan Junli of Microsoft Research,

• 2:09Nathan Labenz?

  have created a small natural language dataset called Tiny Stories, which they designed to reflect the full richness of natural language, while still being small and conceptually simple enough to support research with modest compute budgets. They did this by using GPT-4 to systematically create 1 million children's stories, using only words that an advanced three-year-old could be expected to know. Dataset in hand, they then began to explore a number of aspects of language model performance, behavior, and mechanism by training a series of models that range in size from just 1 million to a maximum of 33 million parameters, still just 2% the size of GPT-2.

• 2:53Nathan Labenz?

  They then used these small models to explore the development of language model reasoning abilities, identifying so-called logical primitives, beginning with a basic understanding of grammar, followed by the learning of facts, and then eventually adding certain logical microskills, such as negation and exclusion.

• 3:13Nathan Labenz?

  These findings create the perfect context in which to discuss the tricky and often controversial topic of emergence, as well as to compare and contrast how large language models learn with how human children learn, and to explain how the differences that we see across language models and children do in fact make some sense, given the different incentive structures in play in each case. They also did some great interpretability work in this paper, and I really relished the chance to get into all three areas that they explore. First, they look at the trade-offs between the number of layers in a transformer, which to a large extent governs the number of logical leaps that a model can make, with, on the other hand, the width of a layer,

• 3:56Nathan Labenz?

  which seems to determine how many facts the model can store.

• 4:01Nathan Labenz?

  They also identify attention heads with distinct roles, including distance heads, which simply reflect the distance between tokens, and which look almost exactly like the alibi scheme, which is now powering long context models such as Claude 100K and the recent Mosaic ML 65K release. And then, on the other hand, semantic attention heads, which focus on meaning. That there should exist such completely different attention heads within a single model, and that an alibi scheme should emerge in the wild, is really, to me, mind-blowing.

• 4:35Nathan Labenz?

  Finally, they examine the role of individual neurons, finding that many of their small model neurons do in fact correspond to human interpretable concepts. We close the conversation by zooming out and discussing why small models are more interpretable than large models, the challenges inherent in attempting to extend this work to larger scale models, and why controlling language models might end up being more like horseback riding than microbiology.

• 5:01Nathan Labenz?

  Throughout this conversation, I was really struck by two things. First, it seems to me that we've only scratched the surface of the potential for curriculum learning approaches. I fully expect that we'll start to see ever more sophisticated approaches, which use specific data sets to layer on specific skills in a strategic, progressive manner, creating highly specialized, small-scale models that can solve specific problems extremely efficiently.

• 5:28Nathan Labenz?

  Second, the value of these toy models for developing understanding really is tremendous. If I could make just one suggestion to listeners, if you want to get the absolute most out of this episode, it would be to visit the Hugging Face website and try playing around with some of the bigger models that they've released. The very biggest are still only 33 million parameters, which means that they can load easily and run quickly right from the Hugging Face model page. If you do that, as I did in preparation for this episode, you will actually have the chance to explore a lot of the concepts, and you can set up your own little experiments to test the reasoning ability of these models. I guarantee that you will come away with a deeper understanding that you will retain for longer and much better.

• 6:10Nathan Labenz?

  And if you do find anything interesting, I would love to hear about it. So please do reach out to me via our email, [email protected], or on Twitter, where you can always DM me, @LeBenz. Now, I hope you enjoy this elucidating conversation with Ronan Eldan and Yuanzhi Li of Microsoft Research. Ronan Eldan and Yuan Zhu Li, welcome to The Cognitive Revolution.

• 6:35Speaker 00

  Thank you so much. We're super happy to be here.

• 6:38Nathan Labenz?

  You guys have just published this paper called Tiny Stories, and I think it's a really fascinating bit of research on multiple levels. So I'm really kind of excited to dive into it with you guys. It touches on a bunch of different themes, including some of the hot-button themes that we'll get to around emergent capabilities and reasoning. And you guys are studying this in a very unique way that makes the problem, I think, more tractable and more approachable, hopefully for our listeners as well. So I'm really excited to introduce this work to them. Maybe just for starters, can you give me a little bit of an introduction to what inspired the Tiny Stories project?

• 7:19Speaker 00

  I guess I'm kind of new to LLMs or deep learning in general. I come from pure math. And when I started looking into architectures, trying to understand what those models are actually doing, how to improve them, et cetera, et cetera. I got very quickly, I got very, very frustrated because it's very easy to come up with ideas. But in order to actually check whether an idea is good, almost always you need to do an experiment that involves a lot of compute. It's just very, very hard to check things.

• 8:05Speaker 00

  You need to, you know, you can either train small models, which basically don't do much in terms of, you know, they don't actually generate text that sounds coherent. You can train like maybe a BERT-sized model, and then it'll do something on some downstream tasks. But whatever it does, doesn't, it doesn't look much like what those LLMs are doing. If you want to really get an LLM experience, you need to do an experiment with a lot of compute that involves, you know, tons of GPUs, et cetera.

• 8:43Speaker 00

  So, you know, for me, it was just a way to address the frustration of not being able to get any insights without, you know, having to do a large experiment.

• 8:57Nathan Labenz?

  And so the main way that you have accomplished that, if I understand correctly, is by kind of narrowing the conceptual space of what both the data set, you know, contains and then obviously what the model is trained to do, right? Like, instead of taking, uh, you know, a small cup out of the whole ocean of mixed up of everything language, you've created a kind of, we're gonna, we're gonna tackle one, you know, very consistent type of input.

• 9:27Speaker 00

  That's fair. I guess we should mention, there have been many attempts to come up with a synthetic or, or non-synthetic, a smaller data set that has all those elements that those large language corpora have, right? So, you know, in language, you have all sorts of elements. You have a lot of facts, you have, so first of all, you have grammar and vocabulary, right? Those are the obvious things you have in language. But then you have facts, you have reasoning that you can infer from those texts, and there's like also many layers of reasoning and, and, you know, there's, I guess there

• 10:12Speaker 00

  are many capabilities involved in being able to parse those data sets. So, you know, our initial motivation was to come up with a data set that has all these qualitative elements, but on the other hand, is just not as massive as those large language corpora. Right, and you know, you Andrew and I had, uh, so first of all, there, as I said, there are many uh synthetic data sets out there. Some of them, I think, reflect in a, in a pretty good way, certain aspects of language,

• 10:54Speaker 00

  such as reasoning, or facts, or grammar, or stuff like that. But we felt like there is no single data set that has all those dimensions together, which are all integrated into something which is not too large, right? And we felt like in order to understand, in order to gain insights about L L M's, we need a data set that has all those elements.

• 11:23Speaker 02

  Yeah, I was just going to add that I, I also came from a D M A background. And, uh, so I was doing theory of machine learning since, uh, maybe seven or eight years ago, when the field just got started. And, yeah, I, I, at that time, like, everyone was doing, uh, research on vision models. And for vision, there's a very nice data set called C far ten, or even M N I S T, I mean, those are very small data sets, they only have like, uh, fifty K images. And when you train on those data set, you can get a pretty high quality image model, and they can do all sort of things. And they reflects what's going on in real large models.

• 12:08Speaker 02

  And, I mean, at that time, uh, doing research or making progress on both theory side and apply side is kind of easy, because they, just training those models only take like one day at most. But when we move to this phase of large language model, or language model in general, the research just becomes so expensive. And I've seen all those blog posts saying that it's impossible to do a P H D now in machine learning without like eight A one hundred. And, I mean, probably only one percent of the P H D student has that amount of compute. So we really want to see whether there is a way to kind of bring the,

• 12:49Speaker 02

  I mean, old, good old days, which are like those C far data set of fast experiment iterations back to the language side. And that's what motivates us to consider this small data set or this simple data set. I think the point is, I mean, there are other kind of synthetic data set or simple data sets that are, as thrown as say, reflecting some aspects of natural languages. But they are not real natural language. They are like just doing simple arithmetic or doing simple like string matching or number manipulation. I mean, they are not real natural language and we want to keep the authentic of natural language,

• 13:32Speaker 02

  but just reduce the overall complexity. So it's still we are studying natural language and not some symbolic manipulation. And still we want the iteration or experiments to be done in a very quick way.

• 13:46Nathan Labenz?

  You created the tiny story, tiny stories data set. I always like to be as concrete as possible, so I'm just gonna read one tiny story uh just straight out of the paper because I think that will help people understand like what this data set, you know, ultimately is. Tom and Jane are friends. One day Jane goes to Tom's house. Tom has a big pot of soup. He wants to share it with Jane. Quote, Jane, do you want some soup? Tom asks. Quote, yes, please. It looks yummy, Jane says. Tom pours some soup into two bowls. He gives one bowl to Jane. Jane takes a spoonful of soup, but then she makes a face. The soup is, now this is just an example presented from the paper, that's the prompt.

• 14:29Nathan Labenz?

  And then you show a completion and you can, you compare and contrast this against other open source models, but I'll just read the twenty eight million parameter version uh that you guys trained. The soup is very bitter. She does not like it. She says, I don't like this soup. It is too bitter. Tom is sorry. He says, I'm sorry, Jane. I didn't know you don't like bitter soup. I will make you something else. He looks around the kitchen and finds some bread and cheese. He puts them on the table and says, here, Jane, you can have some bread and cheese. They are not bitter. They are sweet and yummy. Jane is happy. She says, thank you, Tom. You are a good friend. I like bread and cheese. They are not bitter. So there is our whole tiny story.

• 15:10Nathan Labenz?

  And, you know, I read it almost in like I'm reading to my, you know, four year old or my two year old because it is kind of a children's story. And I understand that that also is kind of part of the motivation. So how did you create this data set? How did you kind of, how do you conceptually, you know, think about those stories you told us a little bit already is kind of having those key elements of the grammar, facts, you know, some amount of like reasoning required.

• 15:37Nathan Labenz?

  But how did you create them? How big is this data set?

• 15:39Speaker 00

  From this motivation to have a good synthetic data set, we should just point out that this is maybe the most natural idea is to rely on, you know, human development, right? It already has the solution for us. Because young children, they are able to speak English somewhat like mostly coherently. I have a daughter, I can testify that not extremely coherently, but somewhat coherently. And, you know, this is, there's already a solution to this coming from like human development, right?

• 16:22Speaker 00

  So all you have to do is just create a data set and make sure that it'll be, basically that any example can be understood by a small child on one hand. And on the other hand, you want it to spend as much as possible of, you know, the knowledge that a small child has. You want it to be as diverse as possible. And we decided, you know, that it makes sense to have this data set somewhat structured. So just the structure of a story, it kind of makes sense because into a story you can,

• 17:06Speaker 00

  inside a story, you can have all those elements combined together, right? Grammar, facts, reasoning, stuff like that. And we just. You know, I think it's a really good time to to to try to create this data set because finally we have those, you know, models GPT three point five and GPT four, which tho- those models can actually understand the instruction. You know, I want a story which is somewhat creative and only has very simple words, right? So GPT four wrote these stories, is that the?

• 17:49Speaker 00

  So, yeah, most of those stories were written by GPT four, some of them by three point five. Three point five is already good enough to write those kind of stories. It's not that great. GPT four is is definitely doing a better job now. It's pretty easy to just write a story, right? Just if I just wanna write a short story, even GPT two can probably write a decent short story. The problem is to actually get a diverse data set that spans, you know, all the vocabulary that you actually wanna span. And if you just ask even GPT four, create a short story and you do it a thousand times

• 18:34Speaker 00

  and you do it with temperature with rather high temperature, let's say, temperature one, which

• 18:43Speaker 00

  kind of gives rise to the mo- the most diversity you can get. Still, about one fifth of the stories will be about children being scared of the slides at the park. We actually, I actually did did this experiment. So it's not very creative. It's if you just tell it to create a story without any other instructions, you're gonna get a very repetitive data set. And so and the whole game is How do you get diversity? How do you make the data sets, you know, not be very repetitive? And here the idea was just to collect a list of

• 19:27Speaker 00

  just a vocabulary of simple words. We have about 2,000 words, which supposedly three-year-olds understand. And then what we do is we just ask GPT-4, okay, here's one random verb, one random noun and one random adjective. Try to combine this into a story in a creative way. We do about a million calls like that. I think we have about 1.5 million stories in the data set. So on one hand, they definitely span all this vocabulary because there's only 2,000 words.

• 20:12Speaker 00

  But on the other hand, you definitely do not span all possible combinations of, you know, words. So you know that you're not gonna, you know, it's not. If you can later create a story with some prescribed combination, you will have demonstrated that the model has some creativity inside it. Yeah, so that's how we created it.

• 20:42Nathan Labenz?

  That's really interesting. Just to do a little bit of math on this. So a million-ish stories, which first of all, that answers the question of why not go use real stories? Because that's a lot of books to scan. So and there may not, I don't know how many children's books there are. You'd have to get your hands on a whole lot to get to a million. So the need for the synthetic data is there. I'm interested in what you were seeing that was like way better about GPT-4 versus 3.5. It sounds like with the Stories that were repetitively about the slides, I would interpret that as maybe like mode collapse, like reflection of kind of, you know, effective RLHF likely. Is that how you would understand that too?

• 21:26Speaker 02

  Uh, I think it's mostly just the model is pre-generating the most likely stories. Because that's what the model is trained. If you don't give the model any content, you will just read all the most likely stories. Because it wants to minimize this language model loss. So potentially just a child scared of the slides is the most common story. Yeah, that exists on the internet, so the model just learn to generate it. Without any given condition, so that's why we want to create some condition to move the model outside. This particular like high probability zoom.

• 22:03Speaker 00

  One difference that we see between GPT three point five and GPT four is. After it, so you give it three words, it needs to combine them somehow in the story. And sometimes it's not that easy, even if I give you three words, I don't know, I think there is an example there. Uh, ancient, uh, thunder and sad or something like that. You wanna combine them in a way that won't look kind of too superficial. You wanna create a story that actually seems fluent. And you don't want like a complete change of topic. In order to, you know, be able to combine the next word.

• 22:49Speaker 00

  And GPT four seems to be able to do this pretty fluently. Um, whereas GPT three point five, sometimes you get stories that don't make so much sense. The words appear there, but they don't appear in a very satisfactory way.

• 23:06Nathan Labenz?

  It it totally makes sense also for you to note, like, yeah, maybe this is just the most common. Story, we don't necessarily have to invoke any like exotic theories of why it keeps talking about slides. But. I wonder if the non RLHF version, you know, if you had access to kind of the base GPT-4 model, if that might have been different. I had the opportunity to red team the GPT-4 early before it had all the safety measures, but it did already have the RLHF and, you know, kind of the instruction following capability. I never saw the like totally base model. I don't think that many people did. They, in the technical report, they said it was not you, you know, not, people weren't sure what to do with it, right? So I think they maybe put that one out to pasture.

• 23:51Nathan Labenz?

  Did you guys try any different versions of GPT-4? Obviously, being, you know, at Microsoft, you, you might have some privileged access to different versions that the rest of us wouldn't have.

• 24:01Speaker 00

  Yeah, so we did, we did have access to an earlier version that, I mean, you know, I, I'm, we're not sure about the exact technical, like, what, what, what the difference is from the model that's now available to the public, but that model had less safety features on it. But as you said, it, it, it may be the same model you had access to. So it did have a certain extent of RLHF, right? Now, I, I think if you take, you know, just the language model trained on the pile without any RLHF and you, you say, you know, create a story such that blah,

• 24:41Speaker 00

  blah, blah, maybe the most plausible completion in terms of a random, you know, entry from the distribution of webpages is, I don't want to do it. Like, write me a story such that, that's the question answer, I don't feel like it. Or, you know, it could be that, that, you know, the net, instead of completing the story, it's just gonna ask, you know, another question without any RLHF, without any alignment that the model has to, to, I mean, the model doesn't know that it's supposed to actually perform your, your instructions,

• 25:23Speaker 00

  right? You know, that's the, the most basic part of it. But I, I think that the RLHF they did on GPT-4 is, is just good enough so that it makes sure to, as accurately as, as it could.

• 25:40Speaker 00

  Actually, you know, satisfy the constraints that you give it. So it almost always combines the words that you ask it to combine into the story. And also, it almost always actually writes a story that only uses simple words.

• 25:57Speaker 02

  Yeah, I think as Ronan point out, the biggest difference of the base like G P T four model compared to this R L H F version is just following instructions. I mean, for base model, you can give it as a beginning of the story, the completion is very good. But if you just say, write me some story that combines certain elements. Then the model has a very hard time understanding the instruction because those things are very rare in the internet. It probably needs some fine tune. So the model understand what does he mean by a instruction. It's not like a conversation, it's an instruction when you ask, write me a story.

• 26:35Nathan Labenz?

  Yeah, makes sense. Uh, yeah, at best, you could maybe do a few shot approach, but then you probably have. I've seen a lot of issues with over indexing on the examples as well. So yeah, I totally, totally get it.

• 26:47Speaker 00

  A few shot approach would probably prime the model into, you know, thinking of specific plots also.

• 26:55Nathan Labenz?

  I'm doing some A I one O one type education at a friend's company right now called Athena. And I was just doing a webinar this morning where I was getting into that with folks saying, you know, if you are going to do a few shot, probably don't do one example in your few shot, at least do two, because otherwise you tend to get this like over. In context, learning on just the one example you gave. So, yeah, I'm with you on on that for sure. So a little bit of math. So there's like two thousand words. There are, you know, let's say that's just to take a round number, um, a thousand each of kind of, you know, the verb, the noun, the adjective. So that in, you know, pure.

• 27:38Nathan Labenz?

  You know, extended expanded form is a billion possible bags of three words roughly speak order magnitude, right? So then you made a million stories. So I just wanted to establish that the space of possibility versus the actual data set that these models were trained on is about a thousand to one ratio. Do I have that roughly right?

• 28:03Speaker 00

  That's true. I think it's pretty accurate. Only in addition to those three words, we also have another another way to add diversity, which is a bunch of features we ask GPT to add to the story, such as a plot twist, a bad ending dialogue. So that adds a little bit more diversity, but one in a thousand is a pretty good like ballpark estimate of the ratio.

• 28:33Nathan Labenz?

  Cool. And then just cost of this, if we were going to pay like retail price for GPT four to write all these million stories, that would be if each one is a 300 tokens, I'll just take a nice round number because that maybe equates it to roughly one cent per it would be like a $10,000 GPT for retail price to generate the data set.

• 29:00Speaker 00

  That's pretty accurate.

• 29:02Nathan Labenz?

  Okay, cool. So then this curriculum concept is, I think, super fascinating. And I this is one of the areas that that had me so intrigued by the paper. You're taking inspiration, you know, obviously, as you said, from human development and, you know, starting with simple words, which definitely makes sense as an approach. I always kind of try to keep in mind as well that like, these things are very alien. And, you know, I do, I'm very intrigued by this curriculum sort of approach, but I wonder.

• 29:34Nathan Labenz?

  What about like more weird curriculums? This is maybe outside of the scope of like this particular research, but I kind of keep waiting for somebody to show up with a like, we trained it first on like pure logic notation. You know, we've seen this a little bit. It's kind of been discussed a lot recently that the code pre-trained models seem to demonstrate better reasoning. You know, once the kind of language part gets added on, you know, and obviously, who knows exactly what that baking recipe looks like? How do you guys think about that? Like, do you expect the same thing? That that somebody's going to pop up with a, hey, we did a pure logic or we did like, you know, just massive amounts of like abstract algebra first

• 30:16Nathan Labenz?

  and kind of taught some sort of, you know, structure that we then were able to layer natural language onto?

• 30:24Speaker 02

  It could definitely help because we, we, based on our previous research on the, I mean, attentions in language models, there are kind of some simple attention mechanisms that the language model may have. The first one is just, it's associating two tokens that are exactly the same. And the second one is after it associate the two tokens that are exactly the same, it also copies the tokens around the first token to the second token. So it's just like us when we read some word, we go back and see what's the previous time that this word appear and what's the surrounding context. And I think just training this head is actually pretty expensive. It requires a lot of training data.

• 31:09Speaker 02

  And something like coding or logic is the perfect way to train those heads. Because for coding, when we define variable, we definitely need to look back like what's the previous definition? Or when we call a function, like we check that function, we see what the function is doing. So it kind of may set the language model into learning those important concepts, like look back or checking the surroundings. And that may serve as a very good warm start for training on other things like simple natural languages. It may make the model learn much faster.

• 31:43Speaker 00

  Yeah, so, so maybe let me, I mean, and another way to say what Yannick said is, Yes, I mean, we do, we do observe this to a certain extent that, you know, maybe coding improves models reasoning. I think at this point, there is no like overwhelming evidence that this is actually the case, but there are some observations. But we are not sure at all that the reason behind it is that, you know, when the model learned how to write code, it actually learned how to reason. It looks like rather the reasons that this, this works are, are, there's, the explanations are just much simpler.

• 32:27Speaker 00

  You just managed to calibrate the exact attention heads that you need. And those attention heads don't have any particular sophistication in them. They might just be able to very accurately look at some relative position to a given token or just compare two tokens in a very precise way. And so, so that means the reason is more like the types of components in your neural network that are required for coding are already there, but those components are, are, are pretty simple.

• 33:08Speaker 00

  It's not like the network has like very sophisticated neural paths that, you know, emerge after the training that actually know how to do reasoning. And for that, we, we actually have a paper we wrote at Microsoft Research about a synthetic task called Lego. So this is a very basic synthetic task that has the core elements of reasoning. And what we observe there is that, so we use a, a transformer based on a BERT architecture. And what we observe is that the pre-trained BERT transformer.

• 33:51Speaker 00

  Basically grasps this reasoning task much faster. The task is is something simple. It's it's basically you get a string that looks like A equals to one, B equals to minus A, C equals to minus B, D equals to C, et cetera, et cetera. And you have to resolve, you know, the values of our variables. And, you know, at first we thought, you know, maybe there is some in some kind of profound sense, the pre-trained BERT model has learned how to reason. And this is why it grasps this task so well.

• 34:31Speaker 00

  But if you dig into it just a little bit, what you realize it's the explanation is just much simpler, much more superficial than that. It's just the pre-training has given rise to some simple attention heads that if if you just initialize the model with those attention heads, then it basically grasps reasoning much faster. This explanation is is is much closer to what actually happens when you train a model to code and then it exhibits better reasoning capabilities.

• 35:09Nathan Labenz?

  So this is maybe a good time to talk about what we mean by reasoning. And I see a ton of confusion out there about this. Maybe you can help us, you know, get a little bit more clarity. I guess one thing that I kind of observe is, you know, of course, people are debating this capability. And it seems like, you know, you've got kind of different standards of evidence or like, you know, people put the burden of proof in different places. To put my cards on the table, you know, I kind of conceive of myself, I call myself these days an AI scout. And I'm really interested in like, what is possible, what can be done, not necessarily

• 35:54Nathan Labenz?

  holding the systems to the standard of that they can do it every time or that they can do it in all cases, you know, or it certainly matters like how adversarially robust they are. But, you know, I wouldn't say, oh, well, we found an example that it failed on, therefore, like it, you know, can't do X if it could do, you know, X nine out of 10 times before it got to that kind of crazy example. But how are you guys thinking about reasoning as, you know, something more than a binary, obviously, in the context of this research?

• 36:28Speaker 02

  Initially, we think reasoning as something that is just a subset of consistency. When we generate sentences or when we say things, we need to make sure that they are consistent with what we said before. And there is like the first level of consistency, which is just nearby words. They need to follow some grammar rules and follow some basic semantics. And those are not really reasoning. It's just some simple, like, more like stochastic parrot, where you just do simple pattern matching, just look at the previous, like, couple of words, and just generate one that is consistent. Well, I think what goes to reasoning is when the consistency goes to the next level,

• 37:09Speaker 02

  which is you really need to consist, be consistent with like something very far away from the current token, like something consistent with a general plot of the story. For example, there's a word but, and then you'll need to say something in the opposite order. And those level of consistency, they are the primitives of reasoning. So we do think that anything beyond just a local consistency should be think of some ability that is reasoning.

• 37:37Speaker 00

  The first thing we have to say is that type of reasoning we're thinking about is a very, very basic. It's just some, you know, basic core capability that comes with speaking coherent English. Some people, I guess, still say that, you know, large language models will never be able to reason. I guess they have a very, very different definition of what reasoning means than what we have, right? I mean, what we mean by reasoning is really like, you know, the capacity to just apply some basic logics when you generate text, right?

• 38:23Speaker 00

  And I think maybe to be concrete, we can look at one of the examples in the paper. So if we look at the sentence like Lily likes cats and dogs. She asked her mom for a dog and her mom said no, so instead she asked. And then you do autocomplete. We kind of see it as a hierarchy of capabilities. So some words in this sentence, in order to to complete them, to to know what the next word is, you just need some very, very basic grammatic rules. For example, she asked her mom for, the next word is a.

• 39:08Speaker 00

  For this, you only need to know a little bit of grammar and that's it. Now, the next word after a, she asked her mom for a dog. Well, you know that the, if if you just know grammar, you know that the next word should be some noun. But here you already need to have some, you know, contextual tracking of, you know, what's going on in the text. The relevant nouns here could be, are probably dogs and cats, right? Those are the two objects that were mentioned in the sentence before.

• 39:49Speaker 00

  Now we go to the next sentence and, you know, you you say like, her mom didn't let her have a dog, so she asked for a. And when you try to autocomplete this, now, you know, the most common noun that you've seen so far, also the most, the the proximate one is dog, not cat. Dog already appears twice in the sentence. You know, she likes dogs and cats. She asked for a dog, her mom said no. So our smaller models actually complete this by saying dog.

• 40:31Speaker 00

  And even GPT two XL, which has one point five billion parameters, its most likely completion is still dog. Because it's still at that level where it is, it did resolve that there should be some noun there. And it does know how to look back in the sentence and see that, okay, there's, there are two nouns, dog and cat. But, you know, dog appears more, so it's more likely that, you know, if I just had a dog in the previous sentence or in the just five words before, it's gonna be dog again.

• 41:13Speaker 00

  But on top of that, if you have a very, very basic reasoning capability, then you're supposed to be able to apply elimination and realize that, okay, she can't have a dog. We had the set containing the two objects, dog and cat. But now dog is not allowed, so what's left is cat. You know, we, we thought this is one of the most basic examples of a completion that would require some extent of, of reasoning.

• 41:53Speaker 02

  There's always this notion, I mean, this intertwine between reasoning and planning. For example, when we say reasoning, many people would think about mathematical reasoning, like I prove a mathematical theorem. And that's not only reasoning, it's also planning. I need to come up with the correct method. I have the intuition, like what's the next step should be. And for us, the reasoning that we are more interested in is just consistency. Like you should say something that are consistent with what you previously said. And the consistency is not only local, it's global. And that's where kind of we think as the reasoning for natural languages.

• 42:33Speaker 00

  The only thing it, it needs to do is generate text that's consistent with the prompt. Like that's, that's the only objective a language model has to, you know, fulfill. The next word it generates should be as consistent as possible with all of the prompt. In order to achieve this consistency, there are several different levels. So for most words that it generates, the only capability that's actually needed is grammar. Or maybe not for most words, but for many words.

• 43:12Speaker 00

  Just by knowing some grammatical rules, you know, you know that if you have a sentence, "Amy wanted," the next word is probably "to." She wanted to something. And you don't need to know anything beyond that, right? Now, the next level after that, and, and this is again, very vaguely speaking. It's not like there is a very strict hierarchy. But the next level is to have some semantic understanding of what's going on, or just to, to understand what are the relevant nouns, actions, stuff like that.

• 43:50Speaker 00

  Or maybe which action is, could be related to which object, et cetera, et cetera. And, you know, if, if, and I, I guess if you look at, uh, models of size, say, around one billion, it's very, very good up to that level. It almost always gives you a word that is grammatically correct. And also has the, it, it, I, I mean, this word is, is kind of well-related. It works well with, you know, the previous few words that you saw.

• 44:31Speaker 00

  Or it, it fits well with, with the previous few words in the, in the prompt. But the next level after that is sometimes already requires, you know, first order logic or second order logic, et cetera.

• 44:49Nathan Labenz?

  So you're kind of breaking it down into like? Micro skills, you know, I, this is a ridiculous analogy, but I'm kind of thinking, I follow this guy on TikTok who coaches basketball micro skills. And it's amazing how many micro skills there are, you know, involved in being a good basketball player. And, you know, mostly the untrained eye, even among basketball fans, like, can't really enumerate them. But this guy has enumerated them and now he's teaching them, you know, one by very small one. And so it's, you know, maybe similarly where you, you wouldn't say like, this person can like fully play basketball. That probably doesn't even like make a lot of sense or sounds, it already sounds strange.

• 45:32Nathan Labenz?

  Uh, but you wouldn't say if there's any missing micro skill that they can't play basketball. You have some sort of continuum there that's like, uh, you know, that people can be better and worse at playing basketball. People can certainly be better and worse at reasoning. And language models too can be better and worse at reasoning. And that probably seems to map on to some sort of hierarchy of micro skills that it either has or it doesn't have, or it's like, you know, maybe in the process of grokking, you know, at any given point in training. So that leads me to the other, you know, big, bold vocabulary word that I want to dig in on a little bit, which is emergence.

• 46:14Nathan Labenz?

  Again, tons of confusion, tons of different, um, you know, meanings out there. I think some people mean like things that surprise us that we didn't necessarily predict. Some people mean like things that happen suddenly. I guess what I kind of think is like, it seems like there is some process where, but I always think back to the, the grokking paper and the Neil Nanda exploration of that, which I'm sure you guys are at least, you know, somewhat familiar with, where there is a phase change from initial memorization to

• 46:50Nathan Labenz?

  A circuit, which, you know, what's so amazing about their work is they actually show this circuit in very like concrete terms. And it's like, and this is the circuit that does this algorithm that allows it to generalize, you know, to the full set, you know, from just the sample data that it was originally trained on. You know, I don't know that we have any circuits here that we could kind of elucidate, but does that feel right to you? Do you feel like there's this sort of process of like memorization sort of being gradually replaced by like concrete circuits that solve particular micro skill challenges? Is that your model of what's going on under the hood here?

• 47:25Speaker 00

  Okay, let's take a step back, okay, like people talk about emergence. I guess we can both agree that, you know, this is a very, this is not a well-defined notion at all. You know, it's not like, it's also, it's not like you see a sudden phase transition from the model not being able to do something to you slightly increase the size and then suddenly it has this, it's really good at some capability it didn't have at all before. Rather, you know, it's a vaguely defined term saying that there are some qualitative capabilities that,

• 48:05Speaker 00

  you know, the model at certain sizes has, whether at smaller sizes, there's almost no trace of these capabilities at all. Like, you know, GPT-2 did not know how to summarize text and suddenly at GPT-3, you have this summarization capability, but the notion is not well-defined at all. On the other hand, we do see that as we increase the size of the model, suddenly you have some certain capabilities you didn't have before. And, you know, in a sense, I think a good analogy is if you compare dogs, monkeys and humans, right?

• 48:49Speaker 00

  You increase the size of the brain, suddenly, you know, humans can do math whereas monkeys cannot. So, you know, it's It's an ability that emerged when you made the neural network larger. Not that I imply, trying to imply at all that the same mechanism explains, you know, both things. We have no idea about that. But, you know, what, what we say here is one of the most important abilities for generative models is to be able to speak coherent English, right? This is an ability that we see emerge

• 49:31Speaker 00

  also in much larger networks trained on, you know, those large language corpora. And I think tiny stories basically gives you a much smaller data set where you can observe this emergence, you know, at much smaller scales of models in, in the sense that, you know, if the model is one million parameters, then it can hardly generate coherent stories. And if you go to 10 million, then almost all stories will be coherent. Same with reasoning, like, you know, one to five million parameter model,

• 50:16Speaker 00

  all of our reasoning prompts fail, whereas for 30 million, almost all of them succeed. Now, as you and you said, all of this basically has to do with keeping coherence with the text. So, so you have the emergency of the ability to generate the next word in a coherent way on different levels of difficulty. So the easiest level of difficulty we could think of is just when you have something that follows from some easy grammatical rules.

• 50:59Speaker 00

  And then, you know, you have, you, you can think of like different level of, levels of difficulty. Sometimes you need to know a certain fact. In order to be able to complete the next word, right? Jack was hungry, so he went looking for some. You to to complete this, you have to know that, you know, to to satisfy the desire for. About like to satisfy hunger, you need some food, right? So sometimes you need to know a fact. So there there is all those core capabilities that are necessary in order to keep consistency along the text.

• 51:44Speaker 00

  And each one of them, we can actually witness its emergence as we increase the size of the model.

• 51:52Nathan Labenz?

  So what then is the theory of what is happening? I have a theory, but I want to hear yours on.

• 51:59Nathan Labenz?

  So you you gave that example a minute ago, right? Girl wants either a cat or dog. Mom says no dog, so it's going to be a cat. But GPT two, you know, much bigger model, whatever. That's like thirty times bigger than your biggest in this. In this research, right? Like you're you kind of max out around thirty million parameters. GPT two is like, what, one point five or something? So it's a lot bigger, fifty times bigger, maybe. It still says dog, which is, you know, every we can all tell that's kind of obviously wrong. You've got these much smaller models that can get that right, you know, in some way, there's this emergent. You know, observed phenomenon that

• 52:44Nathan Labenz?

  it is able to get that, you know, exclusion concept. What do you think is happening there? Is that a micro skill that is like, you know, this sort of exclusion? You know that that's like a little piece of reasoning that is grogged by the small model, but not by the big model. I mean, it seems like there's you could be a really good stochastic parrot, but it feels like there's something there that has like truly kind of settled in to the structure of the network. And maybe that didn't happen with GPT two because the data was just too noisy and it's kind of all over the place. And so it it wasn't able to learn those same things. How am I doing here? Does that does that resonate as likely true or even plausible?

• 53:25Speaker 02

  Yeah, I think that's a valid conjecture. Uh, what we think is for G P T two, because it trained on web data. Or just think of like Wikipedia, where you'll try to minimize the language model loss. The consistency is like the least concern, it's more about just getting the knowledge correct. I mean, you're talking about some object or some person and you want to know his birthday. Or you want to know some specific aspect of that person. I mean, this has nothing to do with natural language, it's more about just the sheer amount of knowledge. That we encounter in the web data or something like Wikipedia. So the model, I mean, because a model, I mean, I think both G P T two and our model.

• 54:10Speaker 02

  They are not large enough to minimize the laws to full extent like G P four. So they have to select some part of the laws that they focus on. And if the data has too many knowledges or too many other nuances. Then the model may just focus on other aspects compared to consistency. Well, here, our tiny story data really pinned down. Like because the language is simple and they, uh, I mean, the vocabularies are simple. The really difficult part is consistency. And that is where the model focus to minimize the training laws. And that's why I think our model, although it's much smaller, it gets better consistency compared to the larger ones.

• 54:52Speaker 00

  Yeah, you can, you can think about it as, you know, when you train a model on the entire pile on, on, you know, those large language corpora. They have much more incentive to learn the clothes, preferred clothing styles of celebrities. Much before they learn how to complete this, you know, sentence with the, the dog and the cat.

• 55:19Speaker 00

  You know, most of, they definitely learned that, you know, Joe Biden is the president of the United States way, way before they learn how to reason. This is a conjecture, of course, I didn't actually test it, but I'm pretty sure that's what happens. You know, you just overload the model with so many facts that appear in so many places. Only once in many words that you generate, this capability of reasoning becomes relevant. So I think it's a really good exercise just to open a random Wikipedia article.

• 56:00Speaker 00

  I did, it's one of my favorite activities since I started thinking about, you know, language models. And just go over the Wikipedia article word by word and just think for every word, what core abilities do you need to use in order to guess, you know, what the next word is? And what you'll see is, I think, definitely for a random Wikipedia article or a random example in the pile, in, you know, some web training set, you will see that only once

• 56:44Speaker 00

  in every 20 to 30 words to predict the next word, you need to use some reasoning capabilities. For one in every three or four words, you need grammar. For most words, you can just kind of guess the next word using, if I just tell you what the, what were the nouns and verbs that appeared in the previous sentences are, without telling you anything about the context or, you know, things which are farther away in the text, you will still be able to guess them. In order, so, so the reasoning is, is kind of a rare capability that you need.

• 57:26Speaker 00

  It only becomes relevant, you know, pretty rarely. And, and therefore you will, you will, like, the capacity of the model will be dedicated to other things much, much before you will get any reasoning capabilities.

• 57:44Speaker 02

  That's also why we think of, why we always kind of see there's an emergence behavior. It's really because the reasoning or those like very rare consistency event, they actually happen very rarely. So even, so only if you minimize the laws to certain extent, you start to learn those rare events. And your model feels like different. For example, just a simple example say, uh, Bob feels uh hungry, but he doesn't like sweet food. So he went to eat, I mean, if the model says went to eat some candy, then we think that this model knows nothing about what he's talking about. Right. But this is only just one word difference out of the ten to twenty words.

• 58:30Speaker 02

  And only if the model get to that extent, it start to learn this consistency. And we feel like the model start to know what it's talking about. But in terms of laws, it's probably we only see like less than ten percent difference. I mean, if you turn into image classification, when I tell you my model get like ninety percent correct and you have a model that get ninety five, I wouldn't consider this as an emergent capability that you get five more percent. But for language model, this five more percent may actually be the emergent uh behavior. And especially for math, like if, if I really want to solve a math problem, there's only like the connection between the two sentences where I need to make sure

• 59:14Speaker 02

  that my proofs are extremely coherent. Well, most of the part, I'm just completing some formulas, write down the results. But it's only this very tiny amount that defines, that that gives the model the emergence capability. So I think the two things are connected, that the model's learning reasoning is like a hard task that it only gets at the very end. And also we see the emergence capability for if you grow the size of the model or if you grow the number of training data.

• 59:45Speaker 00

  Yeah, ma- ma- maybe let me just expand on your example, you know. So, okay, we have this sentence, Bob could have either candy or pizza. Bob doesn't like sweet food, so Bob got some.

• 1:00:00Speaker 00

  And autocomplete, okay? Now, the model, usually if the previous sentence says something about sweet food, if we don't read the entire sentence, the most likely completion is actually candy, not pizza. You have to be, read it in a pretty nuanced way in order to realize that Bob actually doesn't like sweet food, right? Sweet food and candy come together so many, so many times in the training data, right? So, the neural network has to, in quotation marks, choose between using GPT's capacity

• 1:00:42Speaker 00

  in order to be able to resolve this nuance, or to use its capacity in order to know that Joe Biden is the president of the United States, right? You can't have everything together. The model has a finite size and there is some theoretical limit to, you know, the, the amount of things the model can learn. The model will definitely prefer to learn that Joe Biden is president and, you know, many, many other facts because they, they are relevant much, much more frequently.

• 1:01:20Nathan Labenz?

  The curriculum development space is likely to be a huge unlock over the not-too-distant future, right? I mean, you're kind of, it seems like you're probably just kind of scratching the surface here because we've got web-scale data that, you know, is not built for this purpose, obviously. You know, where you're saying reasoning doesn't even, it's not even required that often and so no wonder it kind of emerges late in the game, um, that, you know, maybe pre-training on code, you know, is, is kind of changing that in some interesting ways. But man, intentional design around...

• 1:02:02Nathan Labenz?

  What a gradual, you know, gradually like upstepping curriculum might look like, especially with the ability to, you know, to create the training data synthetically, to, you know, really kind of isolate and bring those, you know, key skills forward. It seems like you could rebalance training data and probably shrink it like a ton and get to a lot of the, you know, just by kind of shifting the balance, right? To get the, these kind of emergent things to be more important relative to just kind of, you know, mind-numbing repetition of, you know, who's the president or whatever. It sounds like, I mean, you're nodding that, it sounds like you, that aligns with your expectations, too.

• 1:02:45Speaker 02

  It's important when we design like a new version of the data, for example, to extend the degree of hours into maybe elementary school or middle school. I think it's very important to balance the amount of knowledge in the data versus the kind of capa- or ability that we want the data set to teach the model. For example, like if you go to elementary school, there's ability to do simple maths or mathematical reasoning or some physics reasoning or do comparison of historical events. I mean, those will take some capa- uh, capacity of the model. Maybe it'll take, it will actually take a very big portion. And the remaining ones, I mean, if your data set has too many knowledges,

• 1:03:28Speaker 02

  then the model may just prefer to use this capability or use this capacity to actually memorize the knowledge instead of really learning those ability. So we want to balance that there's some amount of knowledge that the model must have in order to basic, do basic stuff like maths or do basic physics reasoning. But more importantly, there's, there should be a lot of data that only emphasize on the ability side. Like there's no new knowledge, it's just a bunch of maths training examples or a bunch of simple comparison of some, uh, basic historical events or some simple physical rules and their explanation or different varieties. So that the model can actually focus on the ability part instead of just being screwed

• 1:04:12Speaker 02

  by the vast amount of knowledge. So I think the, right now, the web data, they don't really balance between knowledge and ability training. So that's why training on them is not good for a small model because they need to allocate their capability or capacity to just memorization.

• 1:04:31Speaker 02

  So I think that's basically the criteria for maybe design like better version of synthetic data.

• 1:04:37Speaker 00

  I guess we can just kind of maybe relevant notions here would be the breadth and the depth of the data set or the capabilities of the model. So, you know, the entire web is very, very broad. It's also, so by breadth, I mean, you know, it has a lot of facts. The vocabulary is very large. You need to have a lot of knowledge to capture the data set. And by depth, what I mean is, you know, it has first and second and third order logic that you can infer from learning this data set.

• 1:05:22Speaker 00

  And this is not well established. I mean, there is no, I don't think there is any research that really establishes that there is a trade off between the two. But, you know, it's very, very reasonable to assume that there should be a trade off between breadth and depth when you train the model.

• 1:05:49Speaker 02

  Yeah, I think there's some optimal ratio because without the knowledge, you can't really do reasoning, like you have to have some basic knowledge, let's say candy sweet in order to do reasoning. And when you go to elementary school, I mean, you have to know some basic events, like some basic rules for maths, like one plus one is equal to two, in order to do mathematical reasoning. So there's a balance. You cannot have no knowledge, but you cannot have all the knowledge. So maybe there's some optimal ratio between like breadth and depth.

• 1:06:19Nathan Labenz?

  My head keeps coming back to this same kind of thing where, yeah, we could, should see so much gain from kind of rebalancing the data set and maybe even, you know, starting with some more abstract things, like I could see sort of A or B, not A, therefore, you know, and then do that with like just, you know, all the letters and then start to, you know, introduce kind of these associations and layer that on. Like it sure seems like there is a lot of opportunity there.

• 1:06:49Speaker 02

  Yeah, you're just mentioning the extreme case of only teaching reasoning because there everything is just symbols, there's no knowledge, and it's all about just reasoning. And maybe it's good to combine this with just something that is pure knowledge. And maybe we can get something good by just adjusting their ratio.

• 1:07:08Speaker 00

  Yeah, maybe it's not clear at this point if, you know, a human being, you can, when you teach a human being, you can kind of almost separate the knowledge and the reasoning, right? You can just say, fact A, blah, blah, blah, fact B, blah, blah, blah. Now, here's how to reason. I mean, it has to be a pretty smart human being, but, but in general, you know, we are able to take those two things and then combine them so that in the end, we will have those core reasoning capabilities that we learned using, you know, exercises that only

• 1:07:48Speaker 00

  involve if not A, then if A, A, then B means not B, then not A, right? When we learned, when we studied to the SATs, we have, you know, those rules and, and separately, we have the knowledge of facts and we're able to combine them. And I think it's an important question whether it's even feasible in, in language models to just take those things, separate them into two different modules. And, you know, will the model actually be able to combine those abilities?

• 1:08:27Speaker 00

  My conjecture, by the way, is no, like, as long as, as you don't kind of combine them enough in the data set, the model is not going to be able to infer the way that a human consciously infers the connection. But if we could do that, then this would be, of course, like a very, very powerful technique to train models.

• 1:08:54Nathan Labenz?

  I just did another interview with a couple of guys from Mosaic ML and They talked about training at times on like massive client data sets. Even when they do that, they typically still mix in, you know, kind of the general pre-training, you know, pile or whatever, because otherwise they see catastrophic forgetting. So they have to, you know, kind of keep some mix at all kind of stages of training to avoid that. So I think that would be very consistent, I think, with what you said. Like you, you probably can't do it in strict phases. There's got to be some sort of like mixing strategy throughout the process.

• 1:09:37Speaker 00

  Right, and that's going to be very non-trivial to do. It, it might be feasible, but it's definitely not. And it's, it's not going to happen on its own, just because, you know, what the model cares about is just being able to efficiently autocomplete samples in the data set. If the data set is either knowledge or reasoning, then it has zero incentive to combine. And even if you give it a few examples where this is actually combined, no one is, you know, no one assures you that it'll actually be able to take those two modalities and really use them for the combination.

• 1:10:21Speaker 00

  So I, I feel like this is a science that's really, I mean, we're, we're only beginning to understand how these things work. Hopefully, we'll figure out a way how to actually do it. But I'm not sure, Yu, Yu and you, maybe, I, I don't know any concrete evidence of an example where, where this seems to work.

• 1:10:46Speaker 02

  Uh, right now, we like the kind of concrete evidence that curriculum learning really works, but we believe this should be helpful. But I think overall, as Shou and I say, the model has no incentive to connect like the different phrases where you train things. It just, I mean, when you start a new phase, it's just greedily minimize everything that related to this phase. And it can't just forget everything that he has learned. So it's definitely a very non-trivial task to get this curriculum learning work.

• 1:11:18Nathan Labenz?

  First, you just had a number of kind of interesting empirical observations and you can, you know, address this kind of however you want, but you noted that grammar emerges before consistency and creativity and consistency is, you know, kind of related there to reasoning in our, you know, per your telling earlier. Is that the same thing that I've observed in my kids? I'm kind, you know, my memory, maybe I'm sleep deprived, but I feel like grammar maybe came last for them. Like they definitely have a, you know, a certain consistency, like they want what they want, you know, they like there's, um, if they want ice cream, like, you know, ice cream, they keep the next token is ice cream and it's, they're pretty consistent on that. And, you know,

• 1:12:00Nathan Labenz?

  creativity, I don't know, that's a little trickier, but does this feel like it echoes human development in your mind? Like to me, I'm not sure if it feels that way.

• 1:12:09Speaker 00

  I love this example. Um, so, yeah, I mean, I, I think the, the learning process for children is very, very different that way, right? Children don't get like their incentive is not to say the correct next word. Children, they want ice cream, their incentive, the outcome needs to be, I get ice cream. So if they just say ice cream, ice cream, ice cream many, many times, maybe they don't get the best grade for creativity, but they will likely get the ice cream, right? Depends on how, basically on the self-discipline of parents and whether they are.

• 1:12:52Nathan Labenz?

  Limited in, in this, uh, household.

• 1:12:55Speaker 00

  Yeah. I have a very close experience with, with this exact scenario, but you know, in it, like more seriously, like, you know, when, when, when children produce language, I, I mean, I'm only basing this on my observation. They, they have constant contact with the physical world. They know which entities are involved in the current conversation, right? We are talking about a book that I just. Read, so you know, it's it's very unlikely that the next sentence will be about like a car.

• 1:13:35Speaker 00

  Because you know, they have their in their heads, this entity. Like, no, we're not talking about a car, we're talking about this book, right? Whereas a language model, you know, if it if it just makes one mistake in in one word saying car instead of of book. The loss that it incurs is actually not as big as the loss it would incur for incorrect. Correct grammar, which becomes which is relevant in in almost every word it produces. And not only that, you know, the language model, it it's actually much cheaper, it's much easier.

• 1:14:19Speaker 00

  To have consistent grammar, you only have to kind of be consistent inside the same sentence. For the grammar to be correct, you don't even have to look, you know, past. Five or six words back, whereas for children, you know, for every like. Children have a few entities which are kind of active in their working memory. So this entity could be, let's go, it it could be, I wanna go to the park and have some ice cream. So the two relevant entities are park and ice cream. You're not gonna replace those two entities in the middle of the conversation, that's, you know. What the child cares about at at this point, so.

• 1:15:04Speaker 00

  It's really like uh the the incentive system is different, but also the whole kind of cognitive. Perception around that that that's basically underlying the conversation. Is is pretty different. Yeah, I don't know if it makes sense like that. May maybe a like a good way to think about it for a language model definitely in its first phase of training. It doesn't need to look way back like three, four, five sentences before what it produces.

• 1:15:45Speaker 00

  So so I think a good example is to look at one of the GPT two XL generations that we have in the paper. So we have a story about like a snowman. Two kids are building a snowman and a dog comes and destroys the snowman. And then like you autocomplete to see what happens. And maybe maybe I'll actually read it because I think it emphasize. I think it emphasizes pretty well what I'm trying to say here. So there's those two kids, Sarah and Ben, they built a snowman, a dog comes and destroys the snowman.

• 1:16:31Speaker 00

  And the and now GPT two XL's completion looks something like that. Sarah and Ben cry and look at the snowman in the snow. Ben asks to do something. Sarah tells him to go away. Her hand tightens on the little red button. The snowman is covered in fur. It is red. It is real hair. It is a real snowman. It looks just like the ones they used to make. Ben has a big smile on his face. He looks at the face he has made. He walks over and kisses the eyes. He's happy to be alive. He has done what he was going to do. He has saved the world.

• 1:17:15Speaker 00

  So, OK, what am I actually trying to say here? Except that, you know, GPT two XL has a good sense of humor. What I'm trying to say is if almost every sentence here. The all you need to do in order or or GPT two Excel does in order to generate it is look at the previous sentence. And maybe it knows that there is one entity which is snowman, which is relevant here. Other than that, there is no consistency at all in the completion that goes beyond, you know, looking at two consecutive sentences.

• 1:17:54Speaker 00

  Language models are, if they're too small or in the beginning of training, what they do is they don't, they they don't actually have enough incentive to know the whole context of what's going on. Because more for to complete most words correctly, you just need to have, you know, the context of the current sentence and maybe the one before, and maybe also one or two important entities. Whereas for humans, this is completely different. We have agency, we know what we want when we form the next sentence.

• 1:18:34Speaker 00

  And yeah, we care much less about grammar than about the ice cream.

• 1:18:40Speaker 02

  Yeah, I would make analogy like human child is learning with our RLHF algorithm, just doing reinforcement learning with the parents feedback. And obviously, the parents are very robust to grammar mistakes. So they don't really want to optimize that in order to maximize their reward. They'll probably care more about consistency, the topics, and they want to get the topic correct so they can get the reward. Well, for language model, it's just next word prediction, and the grammar is going to be penalized much more severely compared to the global consistency.

• 1:19:15Nathan Labenz?

  Okay, yeah, I like that. I'm uh, I'm always a little wary of analogies and I'm, I always want to come back and think, what is sneaking into that analogy that I don't want to allow? So I'll bookmark that one, but I, you know, certainly the surface level intuition there uh makes a lot of sense and it's a, it's a good, it's a very clippable analogy as well. How about on the depth versus hidden dimension size of the, which I often just call width, the depth, you know, number of layers versus uh width of a layer? You note some interesting trade offs there, and I didn't really have an intuition necessarily for why that would be, but.

• 1:19:57Nathan Labenz?

  You know, per the paper, you report that the fewer layer models will do better on grammar compared to consistency slash reasoning. And from that, you know, it seems that more layers are important for this kind of reasoning consistency.

• 1:20:17Nathan Labenz?

  How should I think about that? Like, is there a, is there a story that crystallizes why that would be?

• 1:20:24Speaker 00

  None of this is well established. It's all our conjectures that like need to be studied much more. But I think a good way to look at it is so, so depth gives you, tells you about how many times information can percolate between the tokens. So every time you have a global attention layer, like a transformer attention, certain tokens, the information inside certain tokens can percolate into other tokens. So for example, if you have some instruction, create a story with this and these words,

• 1:21:08Speaker 00

  and I want a bad ending as well. And then I type in the beginning of the story and it's, it needs to autocomplete it. These instructions can, in every attention layer, they only have one chance of percolating into the tokens of the story. And sometimes these instructions by themselves are nuanced. So maybe there is an instruction saying the story has a bad ending. It's not enough to know, in order to complete the next word, to know the current sentence and that the story needs to have a bad ending.

• 1:21:50Speaker 00

  You also need to have like the context, like the wider context of what happens in the story. So in order to fulfill these kinds of instructions, you have to basically have the information percolate several times between the tokens. And, and that's, that's also the case for reasoning. If you have first order logic, so if we think about this example of cat and dog, like Alice wanted a cat or a dog, or mother didn't let her have a dog, so she got a cat. How many times does the information have to percolate between the tokens here?

• 1:22:32Speaker 00

  So first, you want to understand that there was a cat and a dog involved. That's one layer of global attention. Then you want to know that she couldn't have a dog. And you really want to, and so you have this set cat and dog. And you want to do cat plus dog minus dog equals to cat. So, so after you know that you have either cat or dog, you need the fact that, you know, she couldn't have a dog to percolate into my token

• 1:23:16Speaker 00

  in order to know that, you know, the only available option is cat. So there are actually two layers of percolation here. You need to know that it's not a dog. So the token not has to go inside the token dog to know that, you know, dog is not allowed. And then those two tokens, not plus dog, have to percolate into the generation in order to know that, okay, I had cat plus dog, but I have to do now minus dog to have a cat. So it's just several, so if you think about it as coding,

• 1:23:57Speaker 00

  you have several, you know, conditionals you have to do and several times that you need to have pointers to information that appears in other places in the text. Okay, this is very, very, I'm putting it in a very vague and non-formal way, but I think, you know, it's a, it's a good initial intuition probably to what happens. Whereas when we talk about facts, so if we, if you have a completion, you know,

• 1:24:36Speaker 00

  I don't know that, okay, like if if I have a completion in a language model that's like the capital of France is, then all I have to do is to is I I just need to have one kind of lookup table with, you know, all the countries and their capitals. I don't have, I I don't need to have many layers of global attention. It's enough to just, you know, take the two tokens, capital and France, put them together and then just have one lookup table saying that capital plus France equals to Paris. Now here the dimension seems to play a much more important role because the bigger the

• 1:25:21Speaker 00

  dimension of the space is, the more entities I can kind of squeeze into this vector space. And, you know, the also the more neurons I have inside my lookup table to tell me, you know, that to to have this list of all those possible facts.

• 1:25:43Nathan Labenz?

  So is another way to say that, that within a single attention block, the attention relationships are not immediately transitive. And so they need multiple iterations of attention in order to create that transitivity. Like if the the current token is looking back at a certain token, but that that token is looking back at a previous token, like we need two rounds of this to move the two hops.

• 1:26:11Speaker 00

  Yeah, exactly. You just, you you first need the not to go into dog to know that it's not dog. And then the not dog needs to go together into this, you know, set that has both cat and dog in it. So that's already two leaps.

• 1:26:30Nathan Labenz?

  Does that also suggest then that sort of? For as many kind of logical leaps as you might need, you need like maybe that many layers, like you can't, you're sort of bounded by, if you have, if you have two layers, you can make maybe two logical leaps. Is that a, a general heuristic that seems sensible?

• 1:26:49Speaker 02

  I think there's a depth and width trade-off. For example, you can simulate two-layer leaps just using one layer, but you have to enumerate like the, all the two possible combinations, which makes your size go from like, say, N to N square. So if you want to be kind of the most size efficient, then you definitely have to go as deep as the number of like logical loops. But if you are not that deep, then you can actually use a wider network to concatenate the two steps into one and just make the intermediate layer like much bigger.

• 1:27:25Speaker 00

  So, so here also, I mean, uh, the, the, your question kind of bursts into an open door in the sense that, uh, the paper we wrote at Microsoft Research about this synthetic reasoning task we call Lego, that's exactly a task where you have multiple leaps of reasoning. And we see a very direct connection between the number of layers that you need and the number of reasoning steps required to complete the task. But maybe somewhat surprisingly,

• 1:28:05Speaker 00

  we see that the model finds like very interesting and sophisticated ways. This is actually not in the paper. This is kind of like a follow-up work. It finds like very sophisticated ways to do multiple leaps of reasoning within a single layer. So, so definitely more layers help, but it's not like it's a strict upper bound for the number of leaps you can, you can do.

• 1:28:35Nathan Labenz?

  I, I could only wish it could be quite that simple. But that's really, really interesting information. I'm learning a lot from this. The interpretability part of this paper is also really interesting. Uh, you kind of break it down into the attention, uh, portion, and then obviously the, you know, the MLP. Uh, you know, neurons portion. And a couple of things jumped out at me. One was in the attention portion. You seem to observe that there's kind of two sorts of attention heads. One really just focuses on the distance relationship between the tokens. And then the other is more semantic.

• 1:29:16Nathan Labenz?

  And the distance one, I was like, holy moly, does that look like the alibi scheme? That has recently come to popularity with, uh, you know, these super long context windows. So I don't know if you guys have had a chance to study that, but quite an uncanny resemblance, right? I mean, you're showing all these attention heads where it's like this one, you know, is just a very tight attention range. And then, you know, they there's different kind of lengths. And that's almost exactly what they cook up as the, you know, kind of substitute for positional embeddings in the alibi, uh, research, at least as far as I understand it. Do you see that same parallel?

• 1:29:58Speaker 02

  Yeah, I think they are definitely doing the same, uh, same thing, which could explain like why alibi is very, I mean, it's very helpful because the positional embedding, uh, positional embedding in alibi is already initialized to do this multi-scale distance-based attention. Well, for absolute positional encoding, like what we use here, the model has to learn to discover this optimal kind of positional-based attention. So just hard code that, uh, positional base or distance-based attention. I think it's a really good choice based on our observation. And also we can, like, the shorter ones are really responsible to just learn the grammars.

• 1:30:40Speaker 02

  And the longer ones, they may just make sure that your content are consistent globally. Or maybe just to grab the associate words. For example, you have an Alice in one sentence and then you have Alice like, uh, five sentences ago. You want to make sure that these two words, you have a chance to put them together.

• 1:30:59Speaker 00

  So, so yeah, let me just point out that.

• 1:31:04Speaker 00

  You know, clearly to complete the next word, you you you you need two things usually. You you wanna know what the proximate words are, what what are the most kind of recent words you saw? And you want to know what are the most important entities in the story are. So these are going to be, you know, words with the relevant semantic meaning uh to what you wanna complete. Now, if I remember correctly, what happens in Alibi is you there is kind of a little bit of a mix of both of them. So you just take every attention head and you make it decay uh.

• 1:31:47Speaker 00

  For every attention head, you just prescribe some scale and the the the strength of attention decays with the distance uh inside the text. Is that is that correct? I if if I'm not mistaken, that's that's what happens there. And we actually see, I mean, one one surprising aspect is we actually see a dichotomy. There are heads that only care about distance and other heads that only care about semantics. And there is hardly a mix between the two.

• 1:32:27Speaker 02

  But we have to say that this is only for one attention block. Uh, we haven't really checked for multiple attention layers, what the transformer would do together. But just for if you just train a network with one attention block, it seems that the network learns to separate the distance-based attention versus the semantic-based attention. Where some heads are just looking at at tokens based on its distance. Some other heads are just looking at tokens based on the semantic similarity.

• 1:32:58Nathan Labenz?

  So is there anything else that you can say? I mean, that's pretty profound in and of itself, that that dichotomy emerges. Um, because you didn't initialize it. I mean, in Alibi, they're they've engineered it that way through, you know, some probably trial and error and heuristics and guesses and whatever. But this is totally just happening on its own. Is there anything else we can say about what you see in the semantic one? When I looked at those, you know, I didn't, no light bulbs went off in my head to kind of interpret those visualizations of the semantic blocks. But anything you would highlight from studying those?

• 1:33:33Speaker 02

  I think the most interesting one we see is just the sematical attention to the main character names. So for example, like there's some had where every token just attend to like the Tom and Lucy, which are the two main characters. I think this is pretty important. I mean, it's just try to identify like what are the person that are involved in the story. So the next time when it generates like a new thing, it's not going to say like Bob, it's going to say something consistent. So I think the semantic has at least what we see in the one transformer block is more about this type of attention where it identify like what are the main objects in the sentence

• 1:34:15Speaker 02

  and just try to make sure that most of the tokens or the relevant tokens attend to those objects. Like when you have the or a, you will attend to like banana. So you know that you want to complete the next word as a banana instead of something completely made up. So I think those sematical attention has they are really useful to just speak consistent English inside the transformers.

• 1:34:40Speaker 00

  Yeah, but let me add that, you know, it's very natural to if to to complete the next word. You want to know what are the relevant characters, what are the relevant entities in the story. But no one expects a priori that you will have such clean attention heads and attention head that exactly attends to the character and a different attention head that exactly, you know, attends to, you know, the objects in the example we gave. It's like a banana and park. A priori, we might expect that it'll all be just a big mess, right? Every attention heads attends to a little bit of everything and, you know, why would

• 1:35:24Speaker 00

  it be interpretable at all? But it's quite surprising with that when the model is small enough, it seems that.

• 1:35:33Speaker 02

  We can actually give meaning to both attention heads and neurons.

• 1:35:38Nathan Labenz?

  So is that, does that kind of fall apart if we add a second layer? Like, does then it just become more messy again? Or what is that, you know, as you start to stack layers, what does that start to look like?

• 1:35:50Speaker 02

  Yeah, I think when your transformers are getting higher or getting deeper or getting larger, it definitely becomes more messy. Because the transformer can simulate, like, I mean, if the transformer is small, you really need to learn those separate modules in order to minimize the loss. But if the transformer has a larger degree of freedom, it has a luxury to, for example, use five attention heads to simulate one or use three layers to do what could be done in one layer. It has no incentive to be like as precise or as kind of conservative as the smaller ones. So it's actually less interpretable. And we also observe that in when we try to interpret the neurons as well.

• 1:36:32Nathan Labenz?

  Perfect bridge then to talk a little bit about the neurons. Maybe just give us a little bit of understanding of like the technique that you used to, you know, I'll, I'll try to summarize it real quick and you tell me where I'm wrong. You run a ton of stories through and you look for what tokens specifically are maximizing the activation of a certain neuron. And then you can kind of print out like, here are the snippets and the individual tokens that maximize the, the activation for this particular neuron. And then holy moly, like, it really looks like there's a pretty coherent concept, you know, as you just kind of scan down that list of things that, you know,

• 1:37:14Nathan Labenz?

  corresponded to high activation on any given neuron.

• 1:37:18Speaker 02

  That's pretty accurate. So you have this, these middle layers in the MLP, which we can think about as neurons. Those are really the coordinates that can either be activated or not. You have the only basically non-linearity in there. And yeah, like, again, just like the attention heads, a priori, it's not.

• 1:37:43Speaker 00

  Not clear at all that they would have any meaning. Those are just, you know, different coordinates of a certain vector space. Like, no one promises you that, you know, the neural network is going to use one particular coordinate for one particular kind of task. And, um, indeed, so, so maybe let me mention that this basically, uh, follows an idea suggested in a twenty fifteen paper by Lee et al. called Visualizing and Understanding Neural Models in NLP, which is just like their idea is just to

• 1:38:27Speaker 00

  look at the tokens which induce the highest activations for every neuron in a certain text. And try to see whether, you know, those to- tokens have a common, uh, role. And when we look at, uh, larger models like, uh, GPT two XL, what we, and we try to look at those tokens, at least the two of us could not find any common meaning. You know, the same neuron is activated sometimes on nouns, sometimes on verbs, sometimes on like, there's, there's just no clear pattern whatsoever. Whereas when we take a small model,

• 1:39:11Speaker 00

  for example, there is one particular neuron that seems to always be activated when the main character of the story is introduced. Um, and, you know, that kind of makes a lot of sense if you think about what the neural networks, network needs to do, I guess. Like, if there was a programmer writing code that tries to autocomplete, you know, stories, there would probably be a function that tries to locate.

• 1:39:48Speaker 00

  The name of the main character because it it's useful in in many, many places. When you autocomplete, in fact, you know, whenever you know that the name of some character should appear. It's a pretty good guess to to think that, you know, this is gonna be the main character. Right? So you have a neuron exactly doing that. And we haven't checked enough to be sure. But there's probably an attention head that that then attends to. You know what this neuron outputs whenever you know that the name of some character should appear.

• 1:40:31Speaker 00

  And when you connect those two together. What you will get is, you know, this mechanism that is able to copy the main character's name to different places along the generation. So, you know, this is a very basic mechanism that you can actually observe inside the neural network. And this doesn't happen in in bigger models, at least not in a way that is, you know, so easy to trace.

• 1:41:03Nathan Labenz?

  I guess my the simple version of it was, I thought maybe it was just like. Maybe there are sort of a bunch of concepts that are easy to identify. Where you can just see like, okay, at a glance, I know what that is. And maybe there's just only so many, you know, like, maybe we only have so you when you have thirty million neurons, you know. Are thirty million parameters, you have however many neurons, you know, maybe that's kind of enough and you can kind of capture those and then you go fifty X and it's like. Well, if you start fishing at random, you know, points in the network there. Maybe you just miss a lot of those, you know, they may exist, but they're just kind of hard to spot because maybe they're sort of sparse, if you will. And then the things that are in between. I mean, I'm really getting out on a limb here, but I was kind of thinking.

• 1:41:45Nathan Labenz?

  Maybe those are sort of analogous a little bit to like the subconscious processing that goes on in our brains where I kind of know on some level that like. Processing is happening.

• 1:41:57Nathan Labenz?

  Even for many concepts, you know, that I don't have like a a clear label for. It's just, you know, there's some sort of churn happening in the brain, but then like only a certain, you know, small set of that kind of rises up to this level of like, you know, what I've called it a conscious concept that I can sort of say, like, I have a label for that and it's like a tidy enough thing. So I guess the two ideas there are, maybe they're just a lot more easy to find in the small network because, you know, you have to have them and, you know, they get packed densely versus a big network. You know, maybe they just are packed more loosely and then those other networks or those other, you know, neurons maybe are just kind of analogous to some stuff that we don't understand very well in our own cognition.

• 1:42:42Speaker 02

  Yeah, it's definitely possible. I think that's an advantage of small language models. They may be more interpretable compared to larger ones because the smaller models can only do basic stuff and only the basic stuffs are probably interpretable. Like the very complicated stuff, for example, how G P T four like write a code that are one thousand lines. I mean, those things is almost impossible to interpret. But how could a small language model keeps a main character consistent? In those basic questions, we can probably understand and there are probably some neurons associated with that. And in G P T four, like out of the, I don't know, maybe ten thousand neurons or even more neurons,

• 1:43:24Speaker 02

  there may also be some neurons that are dedicated to keep the main character consistent. But it's just so hard to find it because it may be in the like twenty five layer neuron, like a nine nine thousand, like seven hundred, whatever. It's just so hard to locate. Well, for smaller models, because it's so small, every neuron must be doing some basic stuff. Because the complicated one, as we said, in the consistency hierarchy or in the loss hierarchy, they only consist of very tiny fraction of the loss. So the main fraction of the loss that contribute to like basic consistency, grammar and things, those are probably the things that are learned by the neurons in the smaller models and they are

• 1:44:08Speaker 02

  more kind of basic and more interpretable.

• 1:44:11Speaker 00

  So you know, there is, there is many ways for the neural network to solve a problem. Like given, given a problem and, and architecture of the neural network, there are many different configurations of the weights that would solve the same problem. Some configurations might be more interpretable to a human and some are just, you know, one big mess. Like every new neuron is little, is doing a little something of every possible task and, you know, they are combined in very, very complicated ways. And then the, the network has no incentive in the loss function not to be a one big mess.

• 1:44:55Speaker 00

  Like most solutions to the same problem are one big mess. You know, this is where the entropy is located, right? And when the model is small, it kind of has no choice. The, the neurons have no choice but to align with meaningful tasks because, you know, the neurons are where you have the nonlinearities and you just don't have enough of them for the one big mess type of solution. Somehow, the most efficient solution is the one that is not completely messy.

• 1:45:35Speaker 00

  And if you have a large network, then, you know, it'll just kind of find a way to do it that does not align with the coordinate structure of the, of the neurons, whereas when the model is small, you just have no choice. So interpretability appears as a, as kind of a side effect.

• 1:45:56Speaker 00

  So anything else that we didn't cover? Yeah, so one thing, maybe going to the initial motivation in creating the data set, which is, you know, to have like some, uh, basically said like a small data set, which is a testing ground for ideas in L L Ms. Like, an open question here is, do we even have a reason to expect that behaviors we witness in this compact setting will translate to L L Ms? Right? We, we don't know the answer to that. Like suppose we find an architecture that works much better

• 1:46:40Speaker 00

  for the Tiny Stories data set. Do we actually have a reason to expect that, you know, this architecture will also be better for L L Ms? Right? So I'm, I'm just saying this as a question. I think it's one of the most relevant questions that, you know, stem from, from this paper. And, and, you know, it, it kind of connects to a more general question, which is, you know, there is all those papers like the, uh, Google Chinchilla paper and the OpenAI Scaling Loss paper, which try to suggest that there might be universal phenomena in L L Ms. There is some, for example, a trade off between width and depth

• 1:47:25Speaker 00

  that is perhaps, I'm not, they don't suggest it explicitly, but there, it, it, but, you know, a natural question that arises, is this universal in the sense that it does not depend on, you know, the exact, uh, mix you take in the data set and the exact architecture and the exact range of sizes you take? So the question here is, are there universal phenomena which will be common to the Tiny Stories data set and to L L Ms being trained on these large corpora?

• 1:48:06Speaker 00

  And yeah, I mean, we, we, maybe let me just say, we, we have just a few indications of some sorts of universality. But at this point, it's completely open and, you know, we, we really hope for the sake of, you know, saving energy and also having, you know, just opening the door for PhD students to actually do LLM research. We hope there is some universality going on so that, you know, you could gain insights, not necessarily on tiny stories, but on any small data set, which would actually be of relevance to LLMs.

• 1:48:49Speaker 02

  Yeah, our future work is mainly just planning to extend the capability of tiny story. If we can create a story that capture like elementary school knowledges. I think this is already a pretty, I mean, it's already a really good data set. If we train a language model, for example, less than one billion, maybe three hundred million parameter. And it's just good that all, everything for elementary school. Or maybe even third grade of elementary school. I think that's already a very good model. I think people will love to interact with it. He knows how to talk, he knows like the basic knowledges. And maybe it's for, I mean, the data set would be diverse enough to capture everything in real language.

• 1:49:34Speaker 02

  Capture every aspect of real language, but just at a, like a down scaled level. And once we have that data set, I think it really opens the door for everyone to do natural language research. Not the ones that has like one hundred, A one hundred in their hands. But the ones with just a laptop GPU, they can train the model in like more two days. And they can gain some interesting observation.

• 1:50:01Speaker 00

  I think what we witness in LLMs is kind of a mathematical miracle going on. And what do I mean by that? You know, you take this algorithm, which is pretty simple. It's, you know, it's gradient descent plus plus. I don't wanna belittle, you know, all the.

• 1:50:21Speaker 00

  Basically, really smart technical contributions that are inside that algorithm. But all in all, you know, it's it's basically gradient descent with an architecture that's very clever, but still it's quite simple. And the miracle is you take, you know, all this this huge training corpus, you fit it to the algorithm, and you don't just get a network that has memorized some text. You get a network that can actually genuinely, you know, create, like synthesize new contact content, show signs of reasoning, understanding, and so on.

• 1:51:03Speaker 00

  And we we think Tiny Stories is just kind of a compact example where you observe the same type of miracle. Of course, it's not as it's not nearly as exciting of what happens in L L M's. But already there, at this size, you you see that there is, you know, some very interesting generalization and emergence going on. And, you know, even if, you know, it doesn't give us a lot of insights about large language

• 1:51:40Speaker 00

  models, this is still kind of a nice playground to try to develop maybe the mathematical foundations necessary to understand why neural networks are able to generalize so well.

• 1:51:59Nathan Labenz?

  So maybe then, just one final question, you know, and I'll encourage people to get in there and try it out. What other interpretability type work have you guys seen that has inspired you that you would recommend that folks in the audience go take a look at as well?

• 1:52:14Speaker 02

  Yeah, I think one of the works that Inspired our research, I mean, it's a work from our group previously, which is called Lego. It's a synthetic reasoning task, which tries to understand what the model is. I mean, try to understand what the attention mechanism of the model is. We identify like several types of different attention has in the in the transformer. Some of them are, as I said, they are just looking at the tokens that are exactly appear before. Like Alice appeared before and Alice is associated with Alice. And there are some other more, I mean, more advanced mechanisms, such as doing deduction or some other stuff.

• 1:52:58Speaker 02

  So I think this work is, I mean, pretty inspiring and it tells us like transformer is at least doing something that is reasonable instead of a pure math. So that's why we also have the interpretability section. We want to look at the attention has and we do see some very good behavior that's corresponding to some aspects of the natural language.

• 1:53:22Speaker 00

  Maybe there is one work I I want to mention also. There is a paper called Transformer Feed Forward Layers Are Key Value Memories. That's another paper I like. That's a paper that tries to interpret what neurons are doing. And I think basically BERT size model transformers. And I mean, they they they are basically able to show that at least some of the neurons have meaningful roles. In general, the theory behind, you know, the interpretability of neural networks is

• 1:54:06Speaker 00

  is it is at its very, very beginning right now. So there are plenty of very, very clever works, but.

• 1:54:17Speaker 00

  It seems just very difficult. So in spite of, you know, really nice works in the literature, I think we are still, you know, light years away of being able to actually understand what's going on inside the, like, the model. And, you know, a priori, there's no reason to assume that we'll ever be able to really understand, right? I mean, we have a very limited understanding of how the human brain works. Like, it's not like we can point to a neuron and say, you know, this neuron has this and this role and, you know, that thought process.

• 1:54:59Speaker 00

  And there's just no reason that we'll be able to ever do it in, in neural networks. And like, there is also no reason to assume that the solution that the neural network finds, that, that solution that gradient descent finds to the problem is not a very, very messy and not interpretable solution. So we'll probably be able to find, to come up with some, you know, kind of basic or, or small examples where, which are partially interpretable. And we might have some insights about big networks.

• 1:55:42Speaker 00

  But, yeah, I, I personally am not very optimistic about, you know, being able to interpret what's happening inside those models to a satisfactory extent that, you know, might lead us to being able to, like, control them and, you know, manipulate them, make sure they have better alignment and so on and so forth.

• 1:56:11Speaker 02

  Yeah, large-scale models interpret we may need to take a different approach. So I think it's impossible to look inside the neural networks and just pin down like the attention is doing something or the neural is doing something. But maybe we have to do a approach that more like our Sparks of AGI paper, where we just talk to the model. And we kind of try to, it's more about interpreting other humans' intention. When we talk to them, I mean, through a sequence of conversations, maybe we can understand what the model like to do and what the model doesn't like to do, or what the model is good at and or what's the typical cases of the model's failure.

• 1:56:52Speaker 02

  And it's more towards like psychology study, but really for robotics, that maybe we need to take that approach for interpretability.

• 1:57:01Speaker 00

  You know, humanity has taken advantage of horseback riding for quite a long time. Now, we have no idea what every neuron inside the horse's brain is doing. We can't really interpret how, you know, we we give some, I don't know how to call it, like command to the horse, like physical cue, and the horse obeys. And it's very, very, very useful. And we can actually rely on like, you know, horseback riding is very reliable. There are very few cases where, you know, the horse has acted unexpectedly in a way that, you know, caused accidents.

• 1:57:46Speaker 00

  Like humanity has profited from that vastly. Uh, maybe, you know, leaving animal rights aside here. And, you know, it works perfectly even without interpretability. And, you know, we we just figured uh figured out ways to align the behavior of the horse with our needs by, you know, taming the horse. We can tame it without understanding the exact process, right? That's going on it on there. And, you know, this is a big success. And I think, you know, we, I think it's just a good analogy, right? I mean,

• 1:58:29Speaker 00

  It's it's kind of horseback riding for the brain. Those L L M's, right? They are just they they give us suddenly we can we can go much faster to much longer distances, even if we don't exactly understand, you know, what the horse is doing like definitely, you know, the Mongols didn't understand much about the biology of the horse. They could still use a horse like in a very reliable way. So, you know, I I just think this is if even though I'm pessimistic about actually

• 1:59:13Speaker 00

  understanding the the inner workings of the neural network, I'm very optimistic about the usefulness and the fact that, you know, we will be able to align it efficiently.

• 1:59:25Nathan Labenz?

  Ronan Eldan and Yuan Junli, thank you for being part of the Cognitive Revolution.

• 1:59:31Speaker 02

  Thank you very much. Yeah, thank you for the invitation. It's really my great pleasure.