Sleep framework: A new way for LLMs to retain knowledge in continual learning

Re:Sleep framework: A new way for LLMs to retain knowledge i... by brabel on Saturday October 31, 2026 @04:54AM (#50314281)

I have a theory that every typo you make distracts the LLM from whatever you are trying to get it to do, so I avoid typos as much as possible. I have no evidence that this is true though.

[ Parent | on HN ]

Re:Sleep framework: A new way for LLMs to retain knowledge i... by derektank on Saturday October 31, 2026 @08:50AM (#50315722)

I’d be willing to believe that. For most of my use cases though, prompt quality doesn’t matter all that much. It’s mostly searching for/in docs

[ Parent | on HN ]