Thank you for your response! In a 5-min article I can't go to the basics, that's why I assumed some previous knowledge.

Parameters are the "synapses" of the neural network. It's where the learning takes place. For each set of examples the neural network is fed, the parameters are updated accordingly to improve classification (or in GPT-3's case, generation of text).

Hope it's more clear now :)

Writing at the intersection of AI, philosophy, and the cognitive sciences | alber.romgar@gmail.com | Get my articles for free: https://mindsoftomorrow.ck.page