Researcher

Melvin Johnson

Melvin Johnson is a Google researcher known for multilingual neural machine translation. He studied computer science at Stanford before joining Google in 2015. In November 2016, he co-authored Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation. The December 2023 Gemini report lists him as lead for instruction tuning, and he is named among the contributors to the Gemini 2.5 report in July 2025. As of October 2026, he works on machine translation and natural language processing at Google.

Now at
Google
Works
47
Citations
11,344
h-index
25

Publication numbers from OpenAlex, as of 15 October 2026.

Timeline

  1. 2015

    Stanford master's, then Google

    He earns a master's in computer science at Stanford, working with Chris Manning, before he joins Google in 2015 to work on machine translation and natural language processing.

  2. November 2016

    Zero-shot multilingual translation

    He co-authors Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation. The paper is submitted to arXiv in November 2016.

  3. December 2023

    Lead for instruction tuning

    He is listed as lead for instruction tuning in the Gemini report. The report is submitted to arXiv in December 2023.

  4. July 2025

    Named among Gemini 2.5 contributors

    He is named among the contributors to the Gemini 2.5 report. The report is submitted to arXiv in July 2025.

  5. July 2026

    An author on Gemma 4

    He is named among the authors of the Gemma 4 technical report. The report is submitted to arXiv in July 2026.

  6. As of October 2026

    Machine translation research at Google

    He works on machine translation and natural language processing at Google.

Sources and credits

Photo: We have no published photo of Melvin Johnson yet. Send us a free-licensed photo.

  1. Melvin Johnson research.google. Checked 15 October 2026.
  2. [1611.04558] Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation arxiv.org. Checked 15 October 2026.
  3. Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation - ACL Anthology aclanthology.org. Checked 15 October 2026.
  4. Gemini: A Family of Highly Capable Multimodal Models arxiv.org. Checked 15 October 2026.
  5. [2312.11805] Gemini: A Family of Highly Capable Multimodal Models arxiv.org. Checked 15 October 2026.
  6. Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities arxiv.org. Checked 15 October 2026.
  7. [2507.06261] Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities arxiv.org. Checked 15 October 2026.
  8. [2607.02770] Gemma 4 Technical Report arxiv.org. Checked 15 October 2026.
  9. Gemma 4 Technical Report arxiv.org. Checked 15 October 2026.