PeopleofSuperintelligence

Researcher

Diana Abagyan

Diana Abagyan is an AI researcher who studies multilingual language models and tokenization. At Columbia, she worked as a peer leader in a computer science program and graduated in 2021. Her research moved from Yoruba dialect NLP to universal tokenizers and Tiny Aya, a multilingual model trained across 70 languages. As of October 2026, she is a Member of Technical Staff at Cohere, specializing in pretraining data for large language models.

Now at
Cohere
Works
6
Citations
3
h-index
1
Badges

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

  1. 2020–2021

    Computer science at Columbia

    As a Columbia College senior, she serves as a peer leader in the Emerging Scholars Program in 2020. She graduates from Columbia College in 2021.

  2. 2024

    Yoruba dialect NLP

    She coauthors Voices Unheard in 2024, which introduces the YORULECT parallel text and speech corpus across three domains and four regional Yoruba dialects. The authors evaluate machine translation, automatic speech recognition, and speech-to-text translation.

  3. 2025

    Research Scholar at Cohere Labs

    She works as a Research Scholar at Cohere Labs in 2025. In August 2025, she presents her research on multilingual tokenizers through the Cohere Labs Regional Africa series.

  4. 2025–2026

    A universal tokenizer

    She releases the One Tokenizer To Rule Them All preprint in 2025 and publishes it in ACL in 2026. The paper proposes a universal tokenizer and reports a language-adaptation win-rate increase of up to 20.2%.

  5. March 2026

    Tiny Aya for 70 languages

    She leads Tiny Aya's pretraining and tokenization efforts and coauthors its technical report, submitted in March 2026. The report describes a 3.35-billion-parameter multilingual model trained on 70 languages.

  6. As of October 2026

    Cohere Member of Technical Staff

    She is a Member of Technical Staff at Cohere and specializes in pretraining data for large language models.

Career

  1. Since 2025CohereResearch Scholar

Education

  1. 2021Columbia College

From the biography.

Most-cited works

  1. Voices Unheard: NLP Resources and Models for Yorùbá Regional Dialects2024 · Conference on Empirical Methods in Natural Language Processing (EMNLP) · 4 citationsWith Noah Smith
  2. One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers2026 · Annual Meeting of the Association for Computational Linguistics (ACL) · 0 citationsWith Sara Hooker
  3. Tiny Aya: Bridging Scale and Multilingual Depth2026 · 0 citationsWith Ahmet Üstün, Beyza Ermis, Daniel D'souza, Joelle Pineau, Phil Blunsom and Sara Hooker

From OpenAlex, as of 19 October 2026.

  1. Google alumni founded or led 87 other organizations, the most among the labs we compareIncludes Cohere. We trace 212 organizations founded or led by AI lab alumni, and 223 career moves between the labs.
  2. The company share of AI researchers’ top papers rose from 2% to 32%Includes Cohere. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 2% in 1975–1979 to 32% in 2020–2024.
  3. Google has the highest combined h-index among AI organizationsIncludes Cohere. Our estimates of combined and median h-index at AI organizations, using current affiliations and OpenAlex records.
  4. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

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

  1. How a Peer-Led Class is Making Computer Science More Equitable | Department of Computer Science, Columbia University cs.columbia.edu. Checked 19 October 2026.
  2. 2021 Columbia College Graduates | Department of Computer Science, Columbia University cs.columbia.edu. Checked 19 October 2026.
  3. Voices Unheard: NLP Resources and Models for Yorùbá Regional Dialects - ACL Anthology aclanthology.org. Checked 19 October 2026.
  4. Open Science Community - Connect Conference labscommunity.cohere.com. Checked 19 October 2026.
  5. Open Science Community - Regional Africa labscommunity.cohere.com. Checked 19 October 2026.
  6. [2506.10766] One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers arxiv.org. Checked 19 October 2026.
  7. One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers - ACL Anthology aclanthology.org. Checked 19 October 2026.
  8. [2603.11510] Tiny Aya: Bridging Scale and Multilingual Depth arxiv.org. Checked 19 October 2026.
  9. Diana Abagyan | SINFO sinfo.org. Checked 19 October 2026.