3 ms·
Yes, LLMs and human brains share at most some faint similarities. Nevertheless, human feats can act as an existence proof of what is possible. Including of wh
by eru 3y ago
Yes, LLMs and human brains share at most some faint similarities.
Nevertheless, human feats can act as an existence proof of what is possible. Including of what might be possible for a neural network.
(I'm not sure whether a large language model necessarily needs to be a neural network in the sense of a bunch of linear transformations interleaved with some simple non-linear activation functions. But for the sake of strengthening your argument, let's assume that we are assuming this restrictive definition of LLM.)