Skip to content

Memory

Definition
AI-generated

Memory in LLM systems is persistent or long-horizon state beyond a single context—conversation summaries, vector stores keyed by embedding, tool results on disk, or explicit banked facts—so agents or chat products can recall user preferences or project constraints across sessions.

Topics
LLM and Agents
Related
AI Agent Context Embedding

Why it matters in GWAS

Assistants that remember prior cohort choices or file paths can speed curation, but memory must be versioned and scoped: stale remembered labels or allele conventions can silently propagate errors across meta-analyses.

Example usage

"We disabled cross-session memory for the public demo so participant-like synthetic records could not be replayed into later chats."

Other meanings (optional)

AI Agent Context Embedding Genome-Wide Association Study (GWAS) Large Language Model (LLM) Retrieval-Augmented Generation (RAG)
Related-term network

Click any node to open that term page.

Node order. The current term is first. Next come every page linked from Related terms, in the same top-to-bottom order as that list. Then new terms appear in two-hop layers: for each of those direct neighbors (in that same order), their related terms are appended if not already listed. A third pass does the same for two-hop nodes. The walk stops when a fixed maximum number of nodes is reached.

References

  • Packer C, et al. (2023). MemGPT: towards LLMs as operating systems. arXiv:2310.08560.
← Meiosis Mendelian Disease →

Last updated (UTC · Git history)