PeopleofSuperintelligence

Researcher

Dan Su

Dan Su is a research scientist at NVIDIA whose work focuses on large language model pretraining and the Nemotron model family. She coauthored CAiRE-COVID, a 2020 question-answering and summarization system, and a 2022 survey of hallucination in natural language generation. She holds a PhD in NLP from the Hong Kong University of Science and Technology. Her NVIDIA work includes the Nemotron-4 340B report, the Nemotron-CC training dataset, Nemotron Nano 2, and Nemotron 3 Ultra.

Now at
NVIDIA
Works
228
Badges

Publication numbers from OpenAlex, as of 19 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2020

    CAiRE-COVID system

    In 2020, she coauthors CAiRE-COVID, a real-time question-answering and multi-document summarization system that wins one of the 10 tasks in the Kaggle COVID-19 Open Research Dataset Challenge.

  2. 2022

    Hallucination in NLG

    In 2022, she coauthors a survey that reviews metrics and mitigation methods for hallucination in natural language generation.

  3. 2024

    The Nemotron-4 report

    In 2024, she coauthors the Nemotron-4 340B technical report, which releases base, instruct, and reward models under NVIDIA's open model license.

  4. 2025

    The Nemotron-CC dataset

    In 2025, she coauthors Nemotron-CC, a 6.3-trillion-token dataset that matches DCLM on MMLU and contains four times more unique real tokens.

  5. August 2025

    NVIDIA Nemotron Nano 2

    In August 2025, she coauthors Nemotron Nano 2; its 9B model reaches up to six times higher inference throughput than similarly sized models in reasoning settings.

  6. June 2026

    NVIDIA Nemotron 3 Ultra

    In June 2026, she coauthors Nemotron 3 Ultra, a 550-billion-parameter mixture-of-experts hybrid Mamba-Attention model with 55 billion active parameters.

  7. As of October 2026

    Research scientist at NVIDIA

    As of October 2026, she is a research scientist at NVIDIA focused on large language model pretraining and holds a PhD in NLP from the Hong Kong University of Science and Technology.

Career

  1. NowNVIDIAResearch scientist

Education

  1. Hong Kong University of Science and TechnologyPhD

From the biography.

  1. 14% of AI career moves in the 2020s went from a university to a companyIncludes Dan Su. Our estimates of how university-to-company career transitions changed by decade, including roles added alongside an earlier job.
  2. Google alumni founded or led 87 other organizations, the most among the labs we compareIncludes NVIDIA. We trace 212 organizations founded or led by AI lab alumni, and 223 career moves between the labs.
  3. 44% of AI career moves go from one university to anotherIncludes NVIDIA. We estimate how 14,272 AI career moves link universities, companies, labs and government research centers, with net flows and the most common paths.
  4. Google has the highest combined h-index among AI organizationsIncludes NVIDIA. Our estimates of combined and median h-index at AI organizations, using current affiliations and OpenAlex records.

Sources and credits

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

  1. CAiRE-COVID: A Question Answering and Query-focused Multi-Document Summarization System for COVID-19 Scholarly Information Management - ACL Anthology aclanthology.org. Checked 19 October 2026.
  2. Survey of Hallucination in Natural Language Generation arxiv.org. Checked 19 October 2026.
  3. Nemotron-4 340B Technical Report arxiv.org. Checked 19 October 2026.
  4. Dan Su - ACL Anthology aclanthology.org. Checked 19 October 2026.
  5. NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model arxiv.org. Checked 19 October 2026.
  6. Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning arxiv.org. Checked 19 October 2026.
  7. Author: Dan Su | NVIDIA Technical Blog developer.nvidia.com. Checked 19 October 2026.