chore(documentation): add context on why 48-vector element

[Warkanlock]

Creating just in case you want to include this information about adding context to the particular 48-element vector explained in the embedding part.

For non-knowledgable people in the field, it might look like a result of the inputs related to the actual architecture of the nano-gpt model.

For me, it took me a while to realize although it will be awesome to add context directly on the visualization

Source: https://github.com/karpathy/minGPT/blob/master/mingpt/model.py

[vercel]

@Warkanlock is attempting to deploy a commit to the Brendan's Team Team on Vercel.

A member of the Team first needs to authorize it.