3 ms·

Yes, it's completely possible that we get to some sort of limit and the scaling laws fall down, loss stops dropping even with more data, parameters, compute. Bu

Yes, it's completely possible that we get to some sort of limit and the scaling laws fall down, loss stops dropping even with more data, parameters, compute. But we have zero reason to believe that is the case, currently, it's only hypothetical. In terms of what we've experimentally seen so far, it just keeps getting better as you scale them up, so to find out if you're right and there's some future roadblock we don't know about we just have to keep scaling these things to see if it works. We haven't even seen loss functions start to level out - the closest argument I've seen to that is people misinterpreting the SGD curve (an artificial tapering we apply to the training process) as a fundamental architecture limitation rather than just an attempt to use resources wisely on a given training run. Basically, training of these models currently only costs a few million dollars, peanuts compared to the budgets of CERN or ITER, and we have zero evidence that we're approaching any sort of ceiling. If we keep scaling, maybe we see some heretofore unpredicted failure that breaks scaling laws like you're predicting. But maybe we don't, and the scaling laws (which have performed incredibly accurately so far) really do hold.

kaba0 3y ago
Having linear scaling up to a point is a far more reasonable guess, than a forever scaling function - as per Occam’s razor. If I have a rubbery string and try lengthening it, it will also show a linear scaling, until it catastrophically fails.
- lucubratory 3y ago
  >Yes, it's completely possible that we get to some sort of limit and the scaling laws fall down, loss stops dropping even with more data, parameters, compute. But we have zero reason to believe that is the case, currently, it's only hypothetical [Edit:- and not really a good hypothetical, because there's no theoretical backing that explains current progress but predicts a future discontinuity]. In terms of what we've experimentally seen so far, it just keeps getting better as you scale them up, so to find out if you're right and there's some future roadblock we don't know about we just have to keep scaling these things to see if it works. Also I have no idea why you think Occam's Razor supports "There will be a discontinuity in this currently accurate observed trend, near enough in the future to be relevant" when the trend is talking about the effects of an algorithm applied to data. By analogy, it's something like claiming Moore's Law won't hold because Occam's Razor suggests at some point there's a physical breakdown, except instead of saying it now you're saying it in 1978, & you're saying it about an algorithmic output rather than a physical thing with an obvious limit (atomic size). Once we see any evidence of all of such a discontinuity, we can adjust what we're doing based on that. Until then, if we want to know whether scaling laws hold we should scale the models.
  - tsimionescu 3y ago
    First of all, there is no theory at all which explains the current progress of LLMs. It's not even a quantifiable progress - it's clearly there, but no one can say with any true precision whether GPT2 to GPT3 to GPT4 is following linear or exponential or whatever else precise function of progress in capabilities. We can clearly see that they are improving, but we have no idea how to accurately quantify this space. Secondly, we can see that there are obvious quantifiable physical increases in the requirements of training and inference for GPT4 compared to GPT3 and to GPT2. So, we can absolutely make a physical argument: at some point, you'll need too much energy to keep scaling up GPTs for it to be feasible, even if capability would keep improving with model size. Thirdly, the GPTs require ever more data to train, and there is a limited amount of text data in the world, of which they are already using a significant percent. So, there is also an argument that we might run out of data to improve even before we run out of physical resources. And making an argument that Moore's law will stop at some point due to physical constraints was just as correct in 1978 as it is in 2024. Predicting when that is is obviously hard - and the same is true for LLM scaling. We may be at the peek, or there may be 100 more generations before we even approach the limits I'm contemplating. But the limits are certainly there.
    - lucubratory 3y ago
      This conversation started with someone saying that despite LLMs having a theoretical basis for how they could emulate human intelligence, they won't actually reach that point because the existing scaling laws won't hold. I am pointing out that we don't know that at all, scaling laws are so far fantastically accurate, and there is no sign of undershooting existing those laws so far. Therefore, it's very premature to declare that LLMs will be unable to reach any given milestone because scaling laws will hit a discontinuity before we get there. Basically: "There isn't even a theoretical basis for how LLMs could achieve human level intelligence" "There is, a neural network trained to predict the next token isn't restricted in the algorithms it can implement internally to achieve that goal, and eventually for loss to be low enough you need to successfully emulate human minds. We have reason to believe loss will continue to go lower because we have very accurate scaling laws which predict that & haven't been wrong so far." "But an LLM won't be able to reach that point because the scaling laws will break down." "Maybe, there's no evidence for that yet but we should scale them up if we want to find out for sure." You are here -> "But in the physical extreme LLMs will break down as an architecture." To which my response is: Yes, of course this specific architecture will break down in the physical extreme. By the time something is building computronium to run the next AI I strongly doubt it will be running matrix multiplication, let alone LLMs specifically. There is no guarantee or reason to believe that we will hit those limits before we are able to achieve human level intelligence or more; the only reason to believe it will happen is an empirically observed trend which could totally break down, so I wouldn't bet the house on it happening, but it's wrong to just rule out the outcome.

kaba0 3y ago

Having linear scaling up to a point is a far more reasonable guess, than a forever scaling function - as per Occam’s razor. If I have a rubbery string and try lengthening it, it will also show a linear scaling, until it catastrophically fails.

lucubratory 3y ago
>Yes, it's completely possible that we get to some sort of limit and the scaling laws fall down, loss stops dropping even with more data, parameters, compute. But we have zero reason to believe that is the case, currently, it's only hypothetical [Edit:- and not really a good hypothetical, because there's no theoretical backing that explains current progress but predicts a future discontinuity]. In terms of what we've experimentally seen so far, it just keeps getting better as you scale them up, so to find out if you're right and there's some future roadblock we don't know about we just have to keep scaling these things to see if it works. Also I have no idea why you think Occam's Razor supports "There will be a discontinuity in this currently accurate observed trend, near enough in the future to be relevant" when the trend is talking about the effects of an algorithm applied to data. By analogy, it's something like claiming Moore's Law won't hold because Occam's Razor suggests at some point there's a physical breakdown, except instead of saying it now you're saying it in 1978, & you're saying it about an algorithmic output rather than a physical thing with an obvious limit (atomic size). Once we see any evidence of all of such a discontinuity, we can adjust what we're doing based on that. Until then, if we want to know whether scaling laws hold we should scale the models.
- tsimionescu 3y ago
  First of all, there is no theory at all which explains the current progress of LLMs. It's not even a quantifiable progress - it's clearly there, but no one can say with any true precision whether GPT2 to GPT3 to GPT4 is following linear or exponential or whatever else precise function of progress in capabilities. We can clearly see that they are improving, but we have no idea how to accurately quantify this space. Secondly, we can see that there are obvious quantifiable physical increases in the requirements of training and inference for GPT4 compared to GPT3 and to GPT2. So, we can absolutely make a physical argument: at some point, you'll need too much energy to keep scaling up GPTs for it to be feasible, even if capability would keep improving with model size. Thirdly, the GPTs require ever more data to train, and there is a limited amount of text data in the world, of which they are already using a significant percent. So, there is also an argument that we might run out of data to improve even before we run out of physical resources. And making an argument that Moore's law will stop at some point due to physical constraints was just as correct in 1978 as it is in 2024. Predicting when that is is obviously hard - and the same is true for LLM scaling. We may be at the peek, or there may be 100 more generations before we even approach the limits I'm contemplating. But the limits are certainly there.
  - lucubratory 3y ago
    This conversation started with someone saying that despite LLMs having a theoretical basis for how they could emulate human intelligence, they won't actually reach that point because the existing scaling laws won't hold. I am pointing out that we don't know that at all, scaling laws are so far fantastically accurate, and there is no sign of undershooting existing those laws so far. Therefore, it's very premature to declare that LLMs will be unable to reach any given milestone because scaling laws will hit a discontinuity before we get there. Basically: "There isn't even a theoretical basis for how LLMs could achieve human level intelligence" "There is, a neural network trained to predict the next token isn't restricted in the algorithms it can implement internally to achieve that goal, and eventually for loss to be low enough you need to successfully emulate human minds. We have reason to believe loss will continue to go lower because we have very accurate scaling laws which predict that & haven't been wrong so far." "But an LLM won't be able to reach that point because the scaling laws will break down." "Maybe, there's no evidence for that yet but we should scale them up if we want to find out for sure." You are here -> "But in the physical extreme LLMs will break down as an architecture." To which my response is: Yes, of course this specific architecture will break down in the physical extreme. By the time something is building computronium to run the next AI I strongly doubt it will be running matrix multiplication, let alone LLMs specifically. There is no guarantee or reason to believe that we will hit those limits before we are able to achieve human level intelligence or more; the only reason to believe it will happen is an empirically observed trend which could totally break down, so I wouldn't bet the house on it happening, but it's wrong to just rule out the outcome.