7 points | by cintusshied 12 hours ago
2 comments
LLMs operate on numbers; LLMs are trained on massive numerical vectors. Therefore, every request is simply a numerical transformation, approximating learned patterns; without proper trainings, their output could be completely irrational.
I mean, define 'reasoning'.
LLMs operate on numbers; LLMs are trained on massive numerical vectors. Therefore, every request is simply a numerical transformation, approximating learned patterns; without proper trainings, their output could be completely irrational.
I mean, define 'reasoning'.