claim accepted goes to: this-edition A all three
The model remembers nothing between requests: every request carries the whole conversation, which is why there is a context window
There is no memory, only a cache that saves work; the model starts from the prompt it receives, so it has to receive all of it, and a long chat costs more every turn. When an assistant 'remembers' you, it is saved notes put back into the prompt.
Decided by. Dinis, in Lenses, roles and loops; the editor records it in the vault
Where it comes from
- Lenses, roles and loops It remembers nothing; Precisely: every request is independent (article)
- Memory is not a spectator sport 'Nothing is remembered by the model between sessions. Everything is remembered by the files.' (newsroom)
What it answers in the 2018 book
- none named
Where it goes
Chapter Lenses, roles and loops, part The frame (drafting); topic How to work with agents: briefs, voice, one shared state.
Other ideas proposed for the same chapter: A prompt is a lens: translating into another language is, to a model, the same as explaining for a teenager or a doctor, A role is a model with a description and a question; an agent is that role on a loop with a goal, Tools changed everything: the model answers with a request, the program runs it, and the result goes into the next request, Long chats are compressed into a lower-resolution summary, so details drop out: work in many focused threads with named agents.
Node idea:the-model-remembers-nothing in graph.json; see it in the graph. Source: admin/content/graph/ideas.json.