News for hackers. Stuff that matters.
Re:OK, I can partly explain the LLM chess weirdness now by hackinthebochs on Friday November 22, 2024 @12:36PM (#42215766)
>The internal model of a LLM is statistical text. Which is linear and fixed.

Not at all. Like seriously, not in the slightest.

Re:OK, I can partly explain the LLM chess weirdness now by skydhash on Friday November 22, 2024 @04:15PM (#42217423)
What does it encode? Images? Scent? Touch? Some higher dimensional qualia?
Re:OK, I can partly explain the LLM chess weirdness now by hackinthebochs on Friday November 22, 2024 @05:27PM (#42217866)
Well, a simple description is that they discover circuits that reproduce the training sequence. It turns out that in the process of this, they recover relevant computational structures that generalize the training sequence. The question of how far they generalize is certainly up for debate. But you can't reasonably deny that they generalize to a certain degree. After all, most sentences they are prompted on are brand new and they mostly respond sensibly.

Their representation of the input is also not linear. Transformers use self-attention which relies on the softmax function, which is non-linear.