PeopleofSuperintelligence

Researcher

Yiyang Nan

Also known as Oliver Nan

Yiyang Nan is a member of technical staff at Cohere who works on multilingual language models and on the evaluation of AI systems. He studied at the University of Michigan and earned an MS in Computer Science at Brown University, advised by Stephen Bach. In 2024, he coauthored the Bonito model for generating instruction-tuning data and the INCLUDE benchmark of regional knowledge in 44 languages. In 2025, he coauthored The Leaderboard Illusion, a study of Chatbot Arena, and the Aya Vision multilingual multimodal models. He joined Cohere as a research scholar in July 2024 and became a member of technical staff in July 2025.

Now at
Cohere
Badges

Timeline

  1. 2019–2021

    Undergraduate at Michigan

    He studies at the University of Michigan, Ann Arbor, from 2019 to 2021, with a major in Data Science, Statistics, and Mathematics of Finance. He works on bioinformatics research with Professor Yuanfang Guan.

  2. 2022–2024

    An MS at Brown University

    He completes an MS in Computer Science at Brown University from 2022 to 2024, advised by Professor Stephen Bach.

  3. February 2024

    Coauthoring the Bonito model

    In February 2024, he coauthors Bonito with Nihal V. Nayak, Avi Trost and Stephen H. Bach, an open-source model that converts unannotated text into task-specific training datasets for instruction tuning.

  4. July 2024–July 2025

    Research Scholar at Cohere

    He joins Cohere as a Research Scholar in July 2024 and holds the post until July 2025.

  5. November 2024

    The INCLUDE benchmark

    In November 2024, he is among the coauthors of INCLUDE, an evaluation suite of 197,243 question-answer pairs from local exams that covers 44 written languages.

  6. April 2025

    The Leaderboard Illusion

    In April 2025, he coauthors The Leaderboard Illusion with Shivalika Singh, Sara Hooker and others, which shows that undisclosed private testing on Chatbot Arena benefits a handful of providers.

  7. May 2025

    Aya Vision multimodal models

    In May 2025, he is the second author, after Saurabh Dash, of Aya Vision, which builds multilingual multimodal models in 8B and 32B sizes. Aya-Vision-32B outperforms models more than twice its size, such as Molmo-72B and LLaMA-3.2-90B-Vision.

  8. Since July 2025

    Technical staff at Cohere

    In July 2025, he becomes a member of technical staff at Cohere. He is a speaker at Cohere Labs Connect 2025, and in February 2026 he coauthors a study of benchmark saturation across 60 language model benchmarks.

Career

  1. Since 2024CohereResearch Scholar, Member of technical staff

Education

  1. 2024Brown UniversityMS in Computer Science
  2. 2021University of Michigan

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Cohere. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.
  2. Among employers with 30+ full-time cases, OpenAI leads primary technical role H-1B pay at $310,000, and Amazon, Google and Microsoft file 57% of all casesIncludes Cohere. Across 440,669 certified US H-1B filings from 44 employers (FY2020–FY2026), OpenAI leads primary technical role base pay at $310,000 and xAI leads overall base pay at $290,000.
  3. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Cohere. Our estimates of where the 943 people shaping AI who worked at Google and Google DeepMind work now: 446 at Google and 452 across 278 other organizations.
  4. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Cohere. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

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

  1. Yiyang (Oliver) Nan nanyyyyyy.github.io. Checked 21 October 2026.
  2. Yiyang Nan | alphaXiv alphaxiv.org. Checked 21 October 2026.
  3. Learning to Generate Instruction Tuning Datasets for Zero-Shot Task Adaptation arxiv.org. Checked 21 October 2026.
  4. INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge arxiv.org. Checked 21 October 2026.
  5. The Leaderboard Illusion arxiv.org. Checked 21 October 2026.
  6. Aya Vision: Advancing the Frontier of Multilingual Multimodality arxiv.org. Checked 21 October 2026.
  7. Open Science Community - Connect Conference labscommunity.cohere.com. Checked 21 October 2026.
  8. When AI Benchmarks Plateau: A Systematic Study of Benchmark Saturation arxiv.org. Checked 21 October 2026.