Topics›Industry and Market Signal›A Working Vocabulary for LLM Memory: Working, Structured, Vector, and Profile Memory
★MediumAct: MediumFull content
A Working Vocabulary for LLM Memory: Working, Structured, Vector, and Profile Memory
ByteByteGo·September 15, 2026
New
Summary
Explainer on LLM memory architecture — trained weights vs. context-window working memory vs. persistent application memory — and memory-extension techniques (sliding window, summarization, structured entity extraction, vector retrieval, long-term profiles, cross-session memory).
Why this signal
General background rather than a novel practice, but the taxonomy is a useful shared vocabulary for designing Thomas Engineering's team-level context layer (focus area #4). Neo starred this.
Original email
ByteByteGo · 9/15/2026
Do LLMs Have the Memory of a Goldfish?