Building LLM Agents by Incorporating Insights from Computer Systems

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Mi, Yapeng, Gao, Zhi, Ma, Xiaojian, Li, Qing
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915231410683904
author Mi, Yapeng
Gao, Zhi
Ma, Xiaojian
Li, Qing
author_facet Mi, Yapeng
Gao, Zhi
Ma, Xiaojian
Li, Qing
contents LLM-driven autonomous agents have emerged as a promising direction in recent years. However, many of these LLM agents are designed empirically or based on intuition, often lacking systematic design principles, which results in diverse agent structures with limited generality and scalability. In this paper, we advocate for building LLM agents by incorporating insights from computer systems. Inspired by the von Neumann architecture, we propose a structured framework for LLM agentic systems, emphasizing modular design and universal principles. Specifically, this paper first provides a comprehensive review of LLM agents from the computer system perspective, then identifies key challenges and future directions inspired by computer system design, and finally explores the learning mechanisms for LLM agents beyond the computer system. The insights gained from this comparative analysis offer a foundation for systematic LLM agent design and advancement.
format Preprint
id arxiv_https___arxiv_org_abs_2504_04485
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Building LLM Agents by Incorporating Insights from Computer Systems
Mi, Yapeng
Gao, Zhi
Ma, Xiaojian
Li, Qing
Computer Vision and Pattern Recognition
LLM-driven autonomous agents have emerged as a promising direction in recent years. However, many of these LLM agents are designed empirically or based on intuition, often lacking systematic design principles, which results in diverse agent structures with limited generality and scalability. In this paper, we advocate for building LLM agents by incorporating insights from computer systems. Inspired by the von Neumann architecture, we propose a structured framework for LLM agentic systems, emphasizing modular design and universal principles. Specifically, this paper first provides a comprehensive review of LLM agents from the computer system perspective, then identifies key challenges and future directions inspired by computer system design, and finally explores the learning mechanisms for LLM agents beyond the computer system. The insights gained from this comparative analysis offer a foundation for systematic LLM agent design and advancement.
title Building LLM Agents by Incorporating Insights from Computer Systems
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.04485