A survey on large language model based autonomous agents

TLDR

Autonomous agents have long been a research focus, yet prior work typically trains them with limited knowledge in isolated environments, diverging from human learning and hindering human‑like decisions; recent large language models trained on vast web knowledge now show human‑level intelligence, spurring research into LLM‑based autonomous agents. This paper presents a comprehensive survey that systematically reviews LLM‑based autonomous agents from a holistic perspective. The survey outlines the construction of LLM‑based autonomous agents, proposes a unified framework, reviews their applications across social science, natural science, and engineering, discusses evaluation strategies, and highlights challenges and future directions.

Abstract

Abstract Autonomous agents have long been a research focus in academic and industry communities. Previous research often focuses on training agents with limited knowledge within isolated environments, which diverges significantly from human learning processes, and makes the agents hard to achieve human-like decisions. Recently, through the acquisition of vast amounts of Web knowledge, large language models (LLMs) have shown potential in human-level intelligence, leading to a surge in research on LLM-based autonomous agents. In this paper, we present a comprehensive survey of these studies, delivering a systematic review of LLM-based autonomous agents from a holistic perspective. We first discuss the construction of LLM-based autonomous agents, proposing a unified framework that encompasses much of previous work. Then, we present a overview of the diverse applications of LLM-based autonomous agents in social science, natural science, and engineering. Finally, we delve into the evaluation strategies commonly used for LLM-based autonomous agents. Based on the previous studies, we also present several challenges and future directions in this field.