Do you have a solution for degradation in accuracy when compiling larger amounts of llm-produced text?
I am also building LLM knowledge/memory systems and I've been surprised how bad LLMs are, even SOTA models, at summarizing non-trivial input batches of text. They get things wrong, distort the underlying meaning or data, etc.
I am also building LLM knowledge/memory systems and I've been surprised how bad LLMs are, even SOTA models, at summarizing non-trivial input batches of text. They get things wrong, distort the underlying meaning or data, etc.