What is fine-tuning?
Teaching a model how to behave — not what is true.
Topic
Assistants that draft, edit and write code — comparisons, pricing and practical workflows.
8 articles
Teaching a model how to behave — not what is true.
Asking for the working, and getting a better answer as a side effect.
Turning meaning into coordinates — the trick that makes semantic search work.
The 2017 architecture underneath essentially every model you have heard of.
The model's working memory — and the reason a long chat starts forgetting things.
The unit models read, count and bill in — and it is not a word.
Four current models, one 1M-token window that is not the same size on all of them, and a tokenizer change nobody announced as a price rise.
Five tools, five different ways of quoting a price — and one of them shows different numbers to different people.