Scholarlink

<p>The article considers the most rapidly developing artificial intelligence technology, large language models. The authors analyze their functionality, providing examples and outlining the potential for their use in various areas of activity. The use of special fine-tuning technologies is shown to enable the creation of numerous neuro-employees on the basis of large language models, which can improve the performance of companies. Fine-tuning adds specialized expertise in a&#xa0;particular field and/or certain functional capabilities to the general intelligence of large language models. The authors describe a&#xa0;pilot project implemented by RCAM-ROSTEST in cooperation with the University of Artificial Intelligence to create a&#xa0;neuro-consultant in the field of legal metrology on the basis of the YandexGPT model. The project results confirmed the practical feasibility and high efficiency of such a&#xa0;neuro-employee. Further development and scaling of the project are possible.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Artificial intelligence-based neuro-consultant in the field of legal metrology

  • 期刊论文

  • 发表日期：

    2025年1月1日

查看全文
  • Alexander Yu. Kuzin,
  • Alexey N. Kroshkin,
  • Ivan A. Obolensky

摘要

The article considers the most rapidly developing artificial intelligence technology, large language models. The authors analyze their functionality, providing examples and outlining the potential for their use in various areas of activity. The use of special fine-tuning technologies is shown to enable the creation of numerous neuro-employees on the basis of large language models, which can improve the performance of companies. Fine-tuning adds specialized expertise in a particular field and/or certain functional capabilities to the general intelligence of large language models. The authors describe a pilot project implemented by RCAM-ROSTEST in cooperation with the University of Artificial Intelligence to create a neuro-consultant in the field of legal metrology on the basis of the YandexGPT model. The project results confirmed the practical feasibility and high efficiency of such a neuro-employee. Further development and scaling of the project are possible.