Research Article
Exploring the Development Status of Agents Based on Large Language Models
@INPROCEEDINGS{10.4108/eai.22-5-2026.2365080, author={Zixi Chen}, title={Exploring the Development Status of Agents Based on Large Language Models}, proceedings={Proceedings of the 4th International Conference on Image, Algorithms, and Artificial Intelligence, ICIAAI 2026, 22-24 May 2026, Singapore, Singapore}, publisher={EAI}, proceedings_a={ICIAAI}, year={2026}, month={8}, keywords={LLMs; Agents; Multi-Agent Systems; Embodied Intelligence; Agentic AI}, doi={10.4108/eai.22-5-2026.2365080} }- Zixi Chen
Year: 2026
Exploring the Development Status of Agents Based on Large Language Models
ICIAAI
EAI
DOI: 10.4108/eai.22-5-2026.2365080
Abstract
The blistering development of Large Language Models (LLMs) is leading to a paradigm shift of artificial intelligence. Of these advancements, agents constructed on the creation of LLM are transforming the models of human-computer interaction and productivity. The given paper is an attempt to analyze the technology of the agents based on the LLM in a systematic way and build a comprehensive framework of cognitions based on micro-level technical systems, up to macro-level patterns of collaboration. It summarizes that the neutralization of the constraints of conventional software can be achieved by the use of the LLM-based agents, which include a set of the brain, planning, memory and the ability to utilize the tools in an unstructured environment. Enterprise-level restructuring, embod intelligence and Agentic AI will be such directions as the development of LLM-based agents in the future.
