[embedding-shape]

> On my team on same the code base you see two different team members producing vastly different results.

And if they use LLMs to assist, does the same thing happen?