PeopleofSuperintelligence

Researcher

Jiachen T. Wang

Also known as Tianhao Wang

Jiachen Wang studies data-centric approaches to trustworthy machine learning. He studied computer science and statistics at Waterloo, earned an MEng at Harvard, and completed a doctorate in electrical and computer engineering at Princeton in May 2026. His work includes robust data valuation, online selection of language-model training batches, and attribution within a single training run; ICLR gave the last work an Outstanding Paper Honorable Mention in 2025. He was named a Rising Star in Data Science in 2024 and received a 2025 Apple Scholars in AI/ML PhD fellowship. After serving as a student researcher at Google Research in 2025 and 2026 and co-organizing ICLR's DATA-FM workshop in 2025, he is a research scientist on ByteDance Seed's Pretrain Data team, a role he began in July 2026.

Now at
ByteDance
Badges

Timeline

  1. 2016–2019

    Computer science at Waterloo

    He earns a B.S. in computer science and statistics at the University of Waterloo in 2019 and works closely with Prof. Florian Kerschbaum.

  2. 2019–2026

    Harvard MEng and Princeton PhD

    He earns an MEng in computational science and engineering at Harvard in 2021, where he works with Prof. Salil Vadhan. He completes a Ph.D. in electrical and computer engineering at Princeton in May 2026 under Prof. Prateek Mittal, and works closely with Prof. Ruoxi Jia at Virginia Tech.

  3. 2023

    Data Banzhaf and KNN-Shapley

    In 2023, he and Ruoxi Jia publish Data Banzhaf, which finds that the Banzhaf value can preserve data rankings when stochastic gradient descent makes model scores noisy. At NeurIPS in 2023, he and his coauthors present Threshold KNN-Shapley, a linear-time, privacy-friendly data valuation method that supports differential privacy.

  4. 2024

    GREATS for LLM data selection

    At NeurIPS in 2024, he coauthors GREATS, an online method that selects data batches during LLM training and improves convergence speed and generalization. He is named a Rising Star in Data Science in 2024.

  5. 2025

    Data Shapley in one training run

    In 2025, ICLR gives Data Shapley in One Training Run an Outstanding Paper Honorable Mention; Jiachen T. Wang is one of its authors. The method attributes data contributions to a target model from a single training run. He also receives a 2025 Apple Scholars in AI/ML PhD fellowship.

  6. 2025–2026

    Google Research and DATA-FM

    He works as a student researcher at Google Research in 2025 and 2026, collaborating with Lin Chen, Mohammadhossein Bateni and Vahab Mirrokni. He co-organizes the second DATA-FM workshop at ICLR in 2025, which addresses data challenges in foundation models.

  7. 2026

    Proxy runs for data curation

    In 2026, he coauthors a study of how small proxy runs can guide data curation for large-scale LLM training. It finds that data-quality conclusions can change with small adjustments to training hyperparameters and proposes reduced learning rates to make proxy evaluations more reliable.

  8. Since July 2026

    Research scientist at ByteDance Seed

    He joins ByteDance Seed's Pretrain Data team as a research scientist in July 2026.

Career

  1. Since 2026ByteDanceResearch scientist
  2. 2025–2026Google ResearchStudent researcher

Education

  1. 2026Princeton UniversityPh.D.
  2. 2021Harvard UniversityMEng
  3. 2019University of WaterlooB.S.

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Jiachen T. Wang. 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. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Jiachen T. Wang. 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.
  3. 16% of AI career moves in the 2020s went from a university to a companyIncludes Jiachen T. Wang. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  4. 23% of people shaping AI who left a university for a company later returned to a universityIncludes Jiachen T. Wang. We estimate that 468 of the 2,011 people shaping AI who left a university for a company (23%) later returned to a university after a median of 5 years, and 163 of 829 (20%) made the reverse round trip.

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 Jiachen T. Wang yet. Send us a free-licensed photo.

  1. Jiachen T. Wang tianhaowang.netlify.app. Checked 21 October 2026.
  2. Data Banzhaf: A Robust Data Valuation Framework for Machine Learning proceedings.mlr.press. Checked 21 October 2026.
  3. Threshold KNN-Shapley: A Linear-Time and Privacy-Friendly Approach to Data Valuation arxiv.org. Checked 21 October 2026.
  4. GREATS: Online Selection of High-Quality Data for LLM Training in Every Iteration proceedings.neurips.cc. Checked 21 October 2026.
  5. Announcing the Outstanding Paper Awards at ICLR 2025 blog.iclr.cc. Checked 21 October 2026.
  6. Data Shapley in One Training Run arxiv.org. Checked 21 October 2026.
  7. DATA-FM Workshop @ ICLR 2025 datafm.github.io. Checked 21 October 2026.
  8. Can Small Training Runs Reliably Guide Data Curation? Rethinking Proxy-Model Practice arxiv.org. Checked 21 October 2026.