long-term memory
In plain wordsMemory that lasts. For an AI agent, it means information saved outside the conversation — facts, past events, learned procedures — that can be looked up and used again days or months later.
- 1
Agents
Information an AI agentAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done. keeps beyond a single context windowcontext window, n.: How much an AI model can read and keep in view at once: the instructions, the conversation so far and any attached documents. Text that does not fit has to be cut, summarized or left out. or session, stored outside the model — in files, databases, vector storesvector database, n.: A database built to store embeddings — lists of numbers that represent meaning — and to quickly find the stored items most similar to a new one. It is the memory behind many AI search tools. or knowledge graphsknowledge graph, n.: A huge web of facts stored as connections between things — 'Paris is the capital of France', 'Marie Curie won a Nobel Prize' — that computers can search, combine and reason over. — and brought back into context when relevant. Following cognitive-science usage, designers distinguish episodicepisodic memory, n.: Memory of particular past events — what happened, when and in what order. AI agents keep records of earlier conversations or attempts so they can recall them and learn from them. memory (records of specific interactions or trajectories), semantic memory (facts about the user, the world or a project) and procedural memory (instructions, skills, code or prompts that encode how to act). Retrieval is usually by similarity search, keywords, recency or explicit file paths. Knowledge absorbed into the weights during training (parametric knowledgeparametric knowledge, n.: What an AI model 'knows' from training, stored in the numbers of its network rather than looked up in documents — like facts you have memorized rather than ones you check in a book.) is a separate, slower-changing store, updated through fine-tuningfine-tuning, n.: Taking an AI model that has already learned a lot and training it a little more on a smaller, focused set of examples so it gets better at one job or behaves a certain way. and continual learningcontinual learning, n.: The ability of an AI to keep learning new things over time from a stream of experience, without forgetting what it already knew and without being retrained from scratch..
We moved user preferences into long-term memory so they survive when the session ends.
- 2
Cognitive science
In psychology, the memory system that retains information for hours to a lifetime, with no known limit on its capacity. It is commonly split into explicit memory, which can be consciously recalled (episodic and semantic), and implicit memory, shown in performance rather than recollection (including procedural skills); unlike working memoryworking memory, n.: The small amount of information held in mind right now while doing a task. For an AI model, that is roughly whatever currently sits in its context window. it is not confined to a handful of items, but recall from it is reconstructive and error-prone.
Origin
The short-term/long-term division goes back to William James's primary and secondary memory (1890) and was formalized in Richard Atkinson and Richard Shiffrin's multi-store model (1968); Endel Tulving separated episodic from semantic memory in 1972. The CoALA framework (2023) carried these categories over to LLM agents, and 'long-term memory' became a chat-assistant product feature from 2024.
Usage note
Not to be confused with long short-term memorylong short-term memory, n.: A type of recurrent network with a separate memory lane and learned 'gates' that decide what to remember, forget and output, letting it keep track of information over long sequences. (LSTM), a recurrent neural-network architecture whose 'memory' is an internal cell state.
Related
- agent memoryagent memory, n.: The ways an AI assistant or agent keeps information — your preferences, past conversations, lessons from earlier tasks — and brings it back later, beyond what fits in a single conversation.
- working memoryworking memory, n.: The small amount of information held in mind right now while doing a task. For an AI model, that is roughly whatever currently sits in its context window.
- episodic memoryepisodic memory, n.: Memory of particular past events — what happened, when and in what order. AI agents keep records of earlier conversations or attempts so they can recall them and learn from them.
- retrieval-augmented generationretrieval-augmented generation, n.: Giving an AI an open-book exam: before answering, it looks up relevant passages in a set of documents and bases its reply on them, often citing where each point came from.
- parametric knowledgeparametric knowledge, n.: What an AI model 'knows' from training, stored in the numbers of its network rather than looked up in documents — like facts you have memorized rather than ones you check in a book.
- MemGPTMemGPT, prop. n.: A 2023 research system that gave a chatbot a computer-style memory hierarchy, letting it move information in and out of its limited working space so it could handle long documents and remember users.
- continual learningcontinual learning, n.: The ability of an AI to keep learning new things over time from a stream of experience, without forgetting what it already knew and without being retrained from scratch.
- working memoryworking memory, n.: The small amount of information held in mind right now while doing a task. For an AI model, that is roughly whatever currently sits in its context window.
- long short-term memorylong short-term memory, n.: A type of recurrent network with a separate memory lane and learned 'gates' that decide what to remember, forget and output, letting it keep track of information over long sequences.
Sources
- Atkinson, R. C. & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In The Psychology of Learning and Motivation, vol. 2. Academic Press.
- Sumers, T. R., Yao, S., Narasimhan, K. & Griffiths, T. L. (2023). Cognitive Architectures for Language Agents. arXiv:2309.02427. arxiv.org/abs/2309.02427v3
Mentioned in
agent memoryagent memory, n.: The ways an AI assistant or agent keeps information — your preferences, past conversations, lessons from earlier tasks — and brings it back later, beyond what fits in a single conversation., vector databasevector database, n.: A database built to store embeddings — lists of numbers that represent meaning — and to quickly find the stored items most similar to a new one. It is the memory behind many AI search tools., working memoryworking memory, n.: The small amount of information held in mind right now while doing a task. For an AI model, that is roughly whatever currently sits in its context window., episodic memoryepisodic memory, n.: Memory of particular past events — what happened, when and in what order. AI agents keep records of earlier conversations or attempts so they can recall them and learn from them., generative agentgenerative agent, n.: A character in a computer simulation, run by a language model, that keeps a diary-like memory, draws lessons from it and schedules its days, so that a group of them can act like a plausible small community., MemGPTMemGPT, prop. n.: A 2023 research system that gave a chatbot a computer-style memory hierarchy, letting it move information in and out of its limited working space so it could handle long documents and remember users.