Stackness
All moves

Retrieval-augmented generation

by Stackness

LLMs

Fetching the relevant documents at question time and putting them in the model's context, instead of relying on what the weights happen to remember.