In Neuroscience, this concept is called "priming"

Our brains, do in fact, retrieve information in a similar fashion to LLM (not exact, but a similar another analogue)

When I say the word "branch" your brain knows it may need information about branches so it will grab for what facts it might need in that moment.

What is a branch? Well it often involves a tree of some sort. A decision tree, a binary tree, a real tree. A real tree? Now it starts thinking about oak, pine, leavess, etc etc. It also filters these out the more information it gleams "Oh you just said birch? Forget about dialogue trees then..."

Funnily enough, your brain gathers information sort of like a tree and also *sort of* like a "trie" in computing.

The difference between us and LLM is an LLM is just a language processor, it predicts a next best word based on all the words it's ever seen but doesn't "think" about it. It runs a set amount of transformations/computations and whatever word equals the next value must be the next word.