Comment on Show HN: Cognita – open-source RAG framework for modular applicationsparentComments−hobs2yAnd really almost always its because the LLMs are really good at summarization and ok at extrapolation and generally lie a lot otherwise.
Comments
And really almost always its because the LLMs are really good at summarization and ok at extrapolation and generally lie a lot otherwise.