Apologies for the late reply, but it turns out I can’t let that sit. Sorry for the rant, but I work in RL and saying “it’s just dice rolls” is insulting to my entire line of work. :(
A probability distribution is not the same as random dice roll. Dice rolls are uniformly and independently random, whereas the probability distributions for LLMs are conditional on the context and the model’s learned parameters. Additionally, all modern LLMs use top K and p sampling–which filters the probability distribution to only high confidence words–so the probability of it choosing to say random garbage is exactly zero.
The issues with LLMs have nothing to do with their sampling from random distributions. That’s just a minor part of their training, and some LLMs don’t even do random sampling since they use tree search. The issues with LLMs are the result of people trying to teach it intelligence using behavior cloning on a corpus of human words and images. Words can’t encode wisdom, only knowledge. Wisdom can only be gained through lived experience.
How well do you think you would perform if you were born into a cave, forced to read a thousand dictionaries in order with no context, and then your only interaction with the outside world was a single question from a single human, and then you died? If you ask me, the LLMs are doing suprisingly well given their “lived experiences”.
“conditional on the context and the model’s learned parameters.” you seem to be under the wrong impression that “random dice roll” == “random dice roll from a uniform distribution”. I didn’t say that. If it outputs a probability distribution, which it does, then you sample it randomly according to that distribution, not a uniform one.
As for your last paragraph: I wasn’t, I didn’t do that, and if that’s all the system can do then people should stop claiming it is even remotely intelligent. Whatever the excuses, the systems aren’t (and won’t be getting) there. If you’re trying to get me to empathize with a couple of matrices, then you’re not going to succeed.
I don’t care if you get offended because someone else doesn’t like your line of work. I think what you do is actively harmful to humanity. I also dislike weapons manufacturers, how they feel about it is irrelevant. You’re no different
you seem to be under the wrong impression that “random dice roll” == “random dice roll from a uniform distribution”.
Almost all dice are uniformly random. Unless you’re using weighted dice? Which, seeing how defensive you get when wrong, might actually make sense 🤔
Knowledge isn’t a competition. Nobody cares about any of this. We will all die and be forgotten. You’re on an internet forum with a bunch of people who have no idea who you are and who could care less about your knowledge of statistics.
Respectfully blocked. I have better things to do with my time.
that’s a massive oversimplification
not really. If the system outputs a probability distribution, then by definition, you’re picking somewhat randomly. So not really a simplification
Apologies for the late reply, but it turns out I can’t let that sit. Sorry for the rant, but I work in RL and saying “it’s just dice rolls” is insulting to my entire line of work. :(
A probability distribution is not the same as random dice roll. Dice rolls are uniformly and independently random, whereas the probability distributions for LLMs are conditional on the context and the model’s learned parameters. Additionally, all modern LLMs use top K and p sampling–which filters the probability distribution to only high confidence words–so the probability of it choosing to say random garbage is exactly zero.
The issues with LLMs have nothing to do with their sampling from random distributions. That’s just a minor part of their training, and some LLMs don’t even do random sampling since they use tree search. The issues with LLMs are the result of people trying to teach it intelligence using behavior cloning on a corpus of human words and images. Words can’t encode wisdom, only knowledge. Wisdom can only be gained through lived experience.
How well do you think you would perform if you were born into a cave, forced to read a thousand dictionaries in order with no context, and then your only interaction with the outside world was a single question from a single human, and then you died? If you ask me, the LLMs are doing suprisingly well given their “lived experiences”.
“conditional on the context and the model’s learned parameters.” you seem to be under the wrong impression that “random dice roll” == “random dice roll from a uniform distribution”. I didn’t say that. If it outputs a probability distribution, which it does, then you sample it randomly according to that distribution, not a uniform one.
As for your last paragraph: I wasn’t, I didn’t do that, and if that’s all the system can do then people should stop claiming it is even remotely intelligent. Whatever the excuses, the systems aren’t (and won’t be getting) there. If you’re trying to get me to empathize with a couple of matrices, then you’re not going to succeed.
I don’t care if you get offended because someone else doesn’t like your line of work. I think what you do is actively harmful to humanity. I also dislike weapons manufacturers, how they feel about it is irrelevant. You’re no different
Almost all dice are uniformly random. Unless you’re using weighted dice? Which, seeing how defensive you get when wrong, might actually make sense 🤔
Knowledge isn’t a competition. Nobody cares about any of this. We will all die and be forgotten. You’re on an internet forum with a bunch of people who have no idea who you are and who could care less about your knowledge of statistics.
Respectfully blocked. I have better things to do with my time.
deleted by creator