Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
| Date | Stars |
|---|---|
| 2026-07-31 | 504 |
| 2026-08-03 | 505 |
| 2026-08-06 | 505 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# A Survey on the Memory Mechanism of Large Language Model based Agents  Large language model (LLM) based agents have recently attracted much attention from the research and industry communities. Compared with original LLMs, LLM-based agents are featured in their self-evolving capability, which is the basis for solving real-world problems that need long-term and complex agent-environment interactions. The key component to support agent-environment interactions is the memory of the agents. While previous studies have proposed many promising memory mechanisms, they are scattered in different papers, and there lacks a systematical review to summarize and compare these works from a holistic perspective, failing to abstract common and effective designing patterns for inspiring future studies. To bridge this gap, in this paper, we propose a comprehensive survey on the memory mechanism of LLM-based agents. In specific, we first discuss ''what is'' and ''why do we need'' the memory in LLM-based agents. Then, we systematically review previous studies on how to design and evaluate the memory module. In addition, we also present many agent applications, where the memory module plays an important role. At last, we analyze the limitations of existing work and show important future directions. **Paper Link:** https://arxiv.org/abs/2404.13501 ## 📆 Updates - [2025/7/2] Our paper has been accepted by ACM Transactions on Information Systems (TOIS) https://dl.acm.org/doi/10.1145/3748302. - [2024/4/21] We release the first version of our survey on arXiv. ## 📋 Table of Contents - [What is the Memory of LLM-based Agent](#-What-is-the-Memory-of-LLM-based-Agent) - [Why do We Need the Memory in LLM-based Agent](#-Why-We-Need-Memory-in-LLM-based-Agent) - [How to Implement the Memory of LLM-based Agent](#-How-to-Implement-the-Memory-of-LLM-based-Agent) - [How to Evaluate the Memory in LLM-based Agent](#-How-to-evaluate-the-memory-in-llm-based-agent) - [Memory-enhanced Agent Applications](#-memory-enhanced-agent-applications) - [Citation](#-citation) - [Acknowledgement](#-acknowledgement) - [Contact Us](#%EF%B8%8F-Contact-Us) ## 🤖 What is the Memory of LLM-based Agent  ## 🤖 Why do We Need the Memory in LLM-based Agent ### Perspective of Cognitive Psychology Cognitive psychology is the scientific study of human mental processes such as attention, language use, memory, perception, problem-solving, creativity, and reasoning. Among these processes, memory is widely recognized as an extremely important one. It is fundamental for humans to learn knowledge by accumulating important information and abstracting high-level concepts, form social norms by remembering cultural values and individual experiences, take reasonable behaviors by imagining the potential positive and negative consequences, and among others. A major goal of LLM-based agents is to replace humans for accomplishing different tasks. To make agents behave like humans, following human's working mechanisms to design the agents is a natural and essential choice. Since memory is important for humans, designing memory modules is also significant for the agents. In addition, cognitive psychology has been studied for a long time. Many effective human memory theories and architectures have been accumulated, which can support more advanced capabilities of the agents. ### Perspective of Self-Evolution To accomplish different practical tasks, agents have to self-evolve in dynamic environments. In the agent-environment interaction process, the memory is key to the following aspects: **(1) Experience accumulation.** An important function of the memory is to remember past error plannings, inappropriate behaviors, or failed experiences, so as to make the agent more effective for handling similar tasks in the future. This is extremely important for enhancing the learning efficiency of the agent in the self-evolving process. **(2) Enviro
Excerpt of 7,016 characters
Read on GitHub
12
3
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:3e6c4465c5a2d4e0, llm:Repository name: 'LLM_Agent_Memory_Survey' — suggests a survey of memory mechanisms for LLM agents (survey/research on agent memory). No topics or README provided.
matched fp:3e6c4465c5a2d4e0, llm:Repository name: 'LLM_Agent_Memory_Survey' — suggests a survey of memory mechanisms for LLM agents (survey/research on agent memory). No topics or README provided.
matched fp:3e6c4465c5a2d4e0, llm:Repository name: 'LLM_Agent_Memory_Survey' — suggests a survey of memory mechanisms for LLM agents (survey/research on agent memory). No topics or README provided.
