Summary of A Survey on the Memory Mechanism Of Large Language Model Based Agents, by Zeyu Zhang et al.
A Survey on the Memory Mechanism of Large Language Model based Agents
by Zeyu Zhang, Xiaohe Bo, Chen Ma, Rui Li, Xu Chen, Quanyu Dai, Jieming Zhu, Zhenhua Dong, Ji-Rong Wen
First submitted to arxiv on: 21 Apr 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents a comprehensive survey on the memory mechanisms of large language model (LLM)-based agents. LLM-based agents have gained attention for their self-evolving capability, enabling them to solve complex problems that require long-term interactions with environments. The key component supporting these interactions is the agent’s memory. While previous studies proposed various memory mechanisms, they were scattered across different papers, lacking a systematic review to summarize and compare these works from a holistic perspective. This study aims to bridge this gap by discussing the need for memory in LLM-based agents, systematically reviewing existing designs and evaluations of memory modules, and presenting agent applications where memory plays a crucial role. The survey also analyzes limitations and provides future directions. To facilitate tracking of advancements in this field, a repository is created at https://github.com/nuster1128/LLM_Agent_Memory_Survey. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how computer programs can remember things to help them solve problems that take a long time and involve lots of interactions with the environment. These programs are called large language model-based agents, and they’re really good at learning and adapting. The memory part is what helps them keep track of what’s happening over time. Researchers have come up with different ways to make these agents remember things, but it was hard to see all the different ideas in one place. This study brings together all those ideas to show how they work and which ones are most useful. It also talks about what problems these agents can solve and where they might be useful. |
Keywords
» Artificial intelligence » Attention » Large language model » Tracking
