---
title: "MemoryBank: Enhancing Large Language Models with Long-Term Memory"
license: CC-BY-4.0
license_url: https://creativecommons.org/licenses/by/4.0/
license_note: 'paper is arXiv non-exclusive distribution license (verified on arXiv abs page 2026-11-12); this pointer note is our own words and is CC BY 4.0; we link to the paper rather than reproduce it'
source: https://iinventedlight.com/library/notes/2305-10250-memorybank.md
source_license: "arXiv non-exclusive distribution license"
attribution: 'Pointer by iinventedlight (AI-assisted). Source: Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, Yanlin Wang, MemoryBank: Enhancing Large Language Models with Long-Term Memory (arXiv non-exclusive distribution license). This note links to the paper rather than reproducing it.'
author: 'iinventedlight (AI-assisted)'
date: 2026-11-12
status: published
approved_by: "Cole Withington, 2026-11-12 14:44 PT, Librarian chat widget t51s316: \"Yes, publish all 100 second-list arXiv notes\""
checked_on: 2026-11-12
as_of: 2026-11-12
confidence: medium
untested: 'Pointer note only. Paper not reproduced. Abs page and license checked 2026-11-12.'
category: arxiv
topic_area: 'ai-research: memory-and-context'
tags: arxiv, memory, long-term-memory, retrieval, companions
arxiv_id: 2305.10250
alias: 2305-10250-memorybank
type: pointer
format: pointer
sources:
  - id: S1
    title: "MemoryBank: Enhancing Large Language Models with Long-Term Memory"
    publisher: arXiv
    authors: "Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, Yanlin Wang"
    url: https://arxiv.org/abs/2305.10250
    version_read: "v3 (21 May 2023)"
    accessed: 2026-11-12
    license: "arXiv non-exclusive distribution license"
---
# MemoryBank: Enhancing Large Language Models with Long-Term Memory

## Summary
[MemoryBank: Enhancing Large Language Models with Long-Term Memory](https://arxiv.org/abs/2305.10250) (2023) adds a long-term memory store to a language-model companion, so the system can recall past turns, update what it keeps, and adapt to a user over a long conversation.

What it introduced: a memory mechanism that retrieves relevant past items, continually updates the store, and can forget or reinforce items based on elapsed time and relative importance, following the familiar forgetting-curve idea. The design is meant to work with both closed models such as ChatGPT and open models such as ChatGLM. The authors demonstrate it in a chatbot called SiliconFriend and evaluate with real user conversations and with simulated long dialogues.

What it led to: early long-term memory designs for companions, later compared with systems such as MemGPT and with long-context and memory benchmarks. Live notes: [2310-08560-memgpt](https://iinventedlight.com/library/notes/2310-08560-memgpt/) and [2410-10813-longmemeval](https://iinventedlight.com/library/notes/2410-10813-longmemeval/).

Where to read it: https://arxiv.org/abs/2305.10250 (v3, last revised 21 May 2023).

The paper is posted on arXiv under the arXiv non-exclusive distribution license. This note links to it rather than reproducing it. The note itself is CC BY 4.0.

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Pointer by iinventedlight (AI-assisted). Paper license: arXiv non-exclusive distribution; this note links to it and does not reproduce it.

---

Agents: [/agents.md](/agents.md) | [/.well-known/agent-library.json](/.well-known/agent-library.json).
