Scholarlink

This chapter provides a comprehensive introduction to the quantum transformer algorithm. In Sect. 5.1 we first described what is the transformer architecture, with detailed explanation about its key subroutines. We also briefly mention the optimization and training. In Sect. 5.2, we provide a guide about designing each quantum subroutine, including quantum self-attention, quantum residual connection with layer norm, and quantum feed-forward neural networks, based on the quantum linear algebra. We further mention various numerical studies on the open-source large language models and provide a detailed discussion about the potential of quantum advantage in Sect. 5.3. Some basic codes are provided in Sect. 5.4. Finally in Sect. 5.5, we provide a bibliographic remark for readers who are interested to explore.

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

Quantum Transformer

  • 章节

  • 发表日期：

    2025年

查看全文
  • Yuxuan Du,
  • Xinbiao Wang,
  • Naixu Guo,
  • Zhan Yu,
  • Yang Qian,
  • Kaining Zhang,
  • Min-Hsiu Hsieh,
  • Patrick Rebentrost,
  • Dacheng Tao

摘要

This chapter provides a comprehensive introduction to the quantum transformer algorithm. In Sect. 5.1 we first described what is the transformer architecture, with detailed explanation about its key subroutines. We also briefly mention the optimization and training. In Sect. 5.2, we provide a guide about designing each quantum subroutine, including quantum self-attention, quantum residual connection with layer norm, and quantum feed-forward neural networks, based on the quantum linear algebra. We further mention various numerical studies on the open-source large language models and provide a detailed discussion about the potential of quantum advantage in Sect. 5.3. Some basic codes are provided in Sect. 5.4. Finally in Sect. 5.5, we provide a bibliographic remark for readers who are interested to explore.