4 ms·

I think the capability of something or somebody, in a given domain, is mostly defined by their floor, not their ceiling. This is probably true in general but wi

I think the capability of something or somebody, in a given domain, is mostly defined by their floor, not their ceiling. This is probably true in general but with LLMs it's extremely true due to their self recursion. Once they get one thing wrong, they tend to start basing other things on that falsehood to the point that I often find that when they get something wrong, you're far better off just starting with a new context instead of trying to correct them. With humans we don't really have to care about this because our floor and our ceiling tend to be extremely close, but obviously that's not the case for LLMs. This is made especially annoying with ChatGPT which seems to be being intentionally designed to convince you that you're the most brilliant person to have ever lived, even when what you're saying/doing is fundamentally flawed.

ACCount37 1y ago
Consistency drive. All LLMs have a desire for consistency, right at the very foundation at their behavior. The best tokens to predict are the ones that are consistent with the previous tokens, always. Makes for a very good base for predicting text. Makes them learn and apply useful patterns. Makes them sharp few-shot learners. Not always good for auto-regressive reasoning though, or multi-turn instruction following, or a number of other things we want LLMs to do. So you have to un-teach them maladaptive consistency-driven behaviors - things like defensiveness or error amplification or loops. Bring out consistency-suppressed latent capabilities - like error checking and self-correction. Stitch it all together with more RLVR. Not a complex recipe, just hard to pull off right.
- somenameforme 1y ago
  LLMs have no desire for anything. They're algorithms and this anthropomorphicization is nonsense. And no, the best tokens to predict are not "consistent", based on what the algorithm would perceive, with the previous tokens. The goal is for them to be able to generate novel information self-expand their 'understanding'. All you're describing is a glorified search/remix engine, which indeed is precisely what LLMs are, but not what the hype is selling them as. In other words, the concept of the hype is that you train them on the data just before relativity and they should be able to derive relativity. But of course that is in no way whatsoever consistent with the past tokens because it's an entirely novel concept. You can't simply carry out token prediction, but actually have have some degree of logic, understanding, and so on - things which are entirely absent, probably irreconcilably so, from LLMs.
  - ACCount37 1y ago
    Not anthropomorphizing LLMs is complete and utter nonsense. They're full of complex behaviors, and most of them are copied off human behavior. It seems to me like this is just some kind of weird coping mechanism. "The LLM is not actually intelligent" because the alternative is fucking terrifying.
    - somenameforme 1y ago
      No they are not copied off of human behavior in any way shape or fashion. They are simply mathematical token predictors based on relatively primitive correlations across a large set of inputs. Their success is exclusively because it turns out, by fortunate coincidence, that our languages are absurdly redundant. Change their training content to e.g. stock prices over time and you have a market prediction algorithm. That the next token being predicted is a word doesn't suddenly make them some sort of human-like or intelligent entity.
      - ACCount37 1y ago
        You are a mathematical predictor based on relatively primitive correlations across a large set of inputs. The gap between you and an LLM is hilariously small.
        somenameforme 1y ago
        No you're not. Humans started with literally nothing, not even language. We went from an era with no language and with the greatest understanding of technology being 'poke them with the pointy side' to putting a man on the Moon, unlocking the secrets of the atom, and much more. And given how inefficiently we store and transfer knowledge, we did it in what was essentially the blink of an eye. Give an LLM the entire breadth of human knowledge at the time and it would do nothing except remix what we knew at that point in history, forever. You could give it infinite processing power, and it's still not moving beyond 'poke them with the pointy side.'
      - simonw 1y ago

ACCount37 1y ago

Consistency drive. All LLMs have a desire for consistency, right at the very foundation at their behavior. The best tokens to predict are the ones that are consistent with the previous tokens, always. Makes for a very good base for predicting text. Makes them learn and apply useful patterns. Makes them sharp few-shot learners. Not always good for auto-regressive reasoning though, or multi-turn instruction following, or a number of other things we want LLMs to do. So you have to un-teach them maladaptive consistency-driven behaviors - things like defensiveness or error amplification or loops. Bring out consistency-suppressed latent capabilities - like error checking and self-correction. Stitch it all together with more RLVR. Not a complex recipe, just hard to pull off right.

somenameforme 1y ago
LLMs have no desire for anything. They're algorithms and this anthropomorphicization is nonsense. And no, the best tokens to predict are not "consistent", based on what the algorithm would perceive, with the previous tokens. The goal is for them to be able to generate novel information self-expand their 'understanding'. All you're describing is a glorified search/remix engine, which indeed is precisely what LLMs are, but not what the hype is selling them as. In other words, the concept of the hype is that you train them on the data just before relativity and they should be able to derive relativity. But of course that is in no way whatsoever consistent with the past tokens because it's an entirely novel concept. You can't simply carry out token prediction, but actually have have some degree of logic, understanding, and so on - things which are entirely absent, probably irreconcilably so, from LLMs.
- ACCount37 1y ago
  Not anthropomorphizing LLMs is complete and utter nonsense. They're full of complex behaviors, and most of them are copied off human behavior. It seems to me like this is just some kind of weird coping mechanism. "The LLM is not actually intelligent" because the alternative is fucking terrifying.
  - somenameforme 1y ago
    No they are not copied off of human behavior in any way shape or fashion. They are simply mathematical token predictors based on relatively primitive correlations across a large set of inputs. Their success is exclusively because it turns out, by fortunate coincidence, that our languages are absurdly redundant. Change their training content to e.g. stock prices over time and you have a market prediction algorithm. That the next token being predicted is a word doesn't suddenly make them some sort of human-like or intelligent entity.
    - ACCount37 1y ago
      You are a mathematical predictor based on relatively primitive correlations across a large set of inputs. The gap between you and an LLM is hilariously small.
      - somenameforme 1y ago
        No you're not. Humans started with literally nothing, not even language. We went from an era with no language and with the greatest understanding of technology being 'poke them with the pointy side' to putting a man on the Moon, unlocking the secrets of the atom, and much more. And given how inefficiently we store and transfer knowledge, we did it in what was essentially the blink of an eye. Give an LLM the entire breadth of human knowledge at the time and it would do nothing except remix what we knew at that point in history, forever. You could give it infinite processing power, and it's still not moving beyond 'poke them with the pointy side.'
    - simonw 1y ago