I have read a book about human brain's work mechanism, then I came out an Idea to code this with PyTorch. I added Curiousity Code to make it better.

Maybe useful in the future robotic technology.

Given an input signal, then found relative or simliar/contrary memory from the historical experience, and enhance/weaken its memory, finally return its thought.

Get interested in new/unseen memory, and store it in a new memory area.

Random selecting the order of resonancer/curiouser to giving the inputed signal, finally return the thought data from resonancer.