2 ms·
But can we say that the LLMs are consciously designed? AFAIK nobody really understands why they work.
by throw-the-towel 3mo ago
But can we say that the LLMs are consciously designed? AFAIK nobody really understands why they work.
- amoss 3mo agoWhether they are or not is irrelevant to the difference- there is still an explicit division between training and inference, with reinforcement of the statistics of the training set.