PeopleofSuperintelligence

Researcher

Kyle Lo

Kyle Lo is an AI researcher whose work spans scientific text, open language models, and vision-language systems. He coauthored SciBERT in 2019, S2ORC in 2020, and Dolma in 2024; Dolma won ACL's Best Resource Paper Award. DrawEduMath appeared among NAACL's 2025 Outstanding Papers, and Molmo and PixMo received CVPR's 2025 Best Paper Honorable Mention. He coauthored OpenScholar in 2026 and worked as a research scientist at Ai2 while co-leading OLMo until March 2026. As of October 2026, his research focuses on large-scale language-model pretraining, data curation, data mixing, benchmark validity, efficient experimentation, and generalization across model scales.

Works
162
Citations
10,782
h-index
35
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2019

    SciBERT for scientific text

    He coauthors SciBERT in 2019, a BERT-based model pretrained on scientific publications to improve downstream scientific NLP tasks.

  2. 2020

    S2ORC for academic papers

    In 2020, he coauthors S2ORC, a corpus of 81.1 million academic papers with structured full text for 8.1 million open-access papers. The corpus supports research and tools for mining academic text.

  3. 2024

    Dolma and an ACL award

    Dolma is the three-trillion-token English corpus he coauthors in 2024 for language-model pretraining. The paper wins ACL's Best Resource Paper Award.

  4. April 2025

    DrawEduMath student math images

    In April 2025, he coauthors DrawEduMath, an English-language dataset of 2,030 images of students’ handwritten responses to K-12 math problems. The team evaluates vision-language models on teacher-written questions and synthetic questions derived from teacher descriptions, and the paper appears among NAACL's Outstanding Papers.

  5. June 2025

    Molmo and PixMo at CVPR

    In June 2025, CVPR recognizes Molmo and PixMo, the vision-language work he coauthors, with a Best Paper Honorable Mention. Molmo is an open family of models with PixMo datasets for image captions, image questions, and pointing.

  6. February 2026

    OpenScholar for literature search

    In February 2026, he coauthors OpenScholar, a retrieval-augmented language model that answers scientific queries using relevant passages from 45 million open-access papers and produces citation-backed responses. The team introduces ScholarQABench, a multi-domain benchmark for scientific literature search.

  7. March 2026

    Leaving Ai2

    Until March 2026, he works as a research scientist at the Allen Institute for AI and co-leads OLMo, its open language-modeling project. He says farewell to Ai2 in March 2026.

  8. As of October 2026

    Language-model pretraining research

    As of October 2026, he is an AI researcher focused on large-scale pretraining of language models, with emphasis on data curation, data mixing, benchmark validity, efficient experimentation, and generalization across model scales.

Career

  1. 2025–2026Worcester Polytechnic Institute
  2. 2020–2025University of California, Berkeley
  3. 2018–2026Allen Institute for AI (Ai2)Research Scientist
  4. 2018–2026University of Washington

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. SciBERT: A Pretrained Language Model for Scientific Text2019 · 3,198 citations
  2. Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks2020 · Annual Meeting of the Association for Computational Linguistics (ACL) · 2,093 citationsWith Noah Smith
  3. BLOOM: A 176B-Parameter Open-Access Multilingual Language Model2023 · 647 citationsWith Alexander Rush, Anna Rogers, Dragomir Radev, Margaret Mitchell, Niklas Muennighoff and Sasha Luccioni
  4. CORD-19: The COVID-19 Open Research Dataset2020 · 537 citationsWith Daniel S. Weld, Dewey Murdick and Oren Etzioni
  5. S2ORC: The Semantic Scholar Open Research Corpus2020 · Annual Meeting of the Association for Computational Linguistics (ACL) · 453 citationsWith Daniel S. Weld

From OpenAlex, as of 21 October 2026.

  1. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Kyle Lo. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes University of Washington. 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.
  3. The median h-index is 48 at universities and 14 at frontier AI labsIncludes University of Washington. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  4. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes University of Washington. 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.

Colleagues

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

Sources and credits

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

  1. SciBERT: A Pretrained Language Model for Scientific Text - ACL Anthology aclanthology.org. Checked 21 October 2026.
  2. S2ORC: The Semantic Scholar Open Research Corpus - ACL Anthology aclanthology.org. Checked 21 October 2026.
  3. Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research - ACL Anthology aclanthology.org. Checked 21 October 2026.
  4. DrawEduMath: Evaluating Vision Language Models with Expert-Annotated Students’ Hand-Drawn Math Images - ACL Anthology aclanthology.org. Checked 21 October 2026.
  5. Announcing the NAACL 2025 Award Winners! 2025.naacl.org. Checked 21 October 2026.
  6. Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models openaccess.thecvf.com. Checked 21 October 2026.
  7. CVPR 2025 Best Papers and Best Demos cvpr.thecvf.com. Checked 21 October 2026.
  8. news | Kyle Lo kyleclo.com. Checked 21 October 2026.
  9. publications | Kyle Lo kyleclo.com. Checked 21 October 2026.
  10. Kyle Lo kyleclo.com. Checked 21 October 2026.