Research Article
Analyzing the Algorithms and Evaluation of Homo Sapiens Artificial Intelligence Agents in Application - A Case Study of Gaming
@INPROCEEDINGS{10.4108/eai.22-5-2026.2365242, author={Yanting Yu}, title={Analyzing the Algorithms and Evaluation of Homo Sapiens Artificial Intelligence Agents in Application - A Case Study of Gaming}, 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={Game Testbeds; AI Agents; Reinforcement Learning; General Intelligence Evaluation; Multi-Agent Systems}, doi={10.4108/eai.22-5-2026.2365242} }- Yanting Yu
Year: 2026
Analyzing the Algorithms and Evaluation of Homo Sapiens Artificial Intelligence Agents in Application - A Case Study of Gaming
ICIAAI
EAI
DOI: 10.4108/eai.22-5-2026.2365242
Abstract
Traditional real-world scenarios are often costly, uncontrollable, and difficult to reproduce, while game environments have gradually become an important testbed for the research and evaluation of AI algorithms. This paper systematically reviews the application of games as complex testbeds in the research of AI agent algorithms, focusing on analyzing the testing value of different types of game environments, the current development status of mainstream AI agent algorithms, and common evaluation metrics and test frameworks.This paper focuses on studying and summarizing the application characteristics of game testbeds in AI agent research, clarifying the differentiated value of different types of game environments for algorithm testing, and clarifying the development trajectory and applicable scenarios of mainstream AI agent algorithms. At the same time, it summarizes the core challenges currently faced in this field and points out that the combination of large language models and multi-agent systems is an important future development direction.
