长春工程学院学报（自然科学版）2025，Vol.26Issue(1)：108-113,6.DOI:10.3969/j.issn.1009-8984.2025.01.018
基于大语言模型的人工智能生成语料判别的研究
A Study of AI Generated Corpus Discriminations Based on Large Language Model
摘要
Abstract
With the widespread application of artificial intelligence technology in the social field,false Chi-nese texts generated by artificial intelligence triggers a series of social problems such as academic integrity.However,existing common detection tools have shown significant shortcomings in addressing such issues.To this end,a solution based on a large language model is proposed,which uses artificial intelligence tech-nology to specifically detect text generated by artificial intelligence.Through systematic verification of four sets of experimental data,it is found that large language models with different parameter scales exhibit sig-nificant performance differences.The experimental results show that this method outperforms existing ci-vilian grade solutions in terms of detection accuracy and cost-effectiveness.Further research reveals that the number of model parameters is not linearly correlated with performance,and specific architecture mod-els can achieve better performance balance at lower parameter levels.关键词
生成式人工智能/AIGC/大语言模型/ChatGPTKey words
Generative Atificial Intelligence/AIGC/large language model/Chat GPT分类
信息技术与安全科学引用本文复制引用
余若岩,张华,李玥曦,桑赫阳,王泽轩..基于大语言模型的人工智能生成语料判别的研究[J].长春工程学院学报（自然科学版）,2025,26(1):108-113,6.基金项目
长春工程学院大学生创新创业训练计划资助项目(S202411437096) （S202411437096）
