PeopleofSuperintelligence

Researcher

Jay Whang

Jay Whang is a research scientist at Google DeepMind who works on Gemini and was a core contributor to Google's Imagen text-to-image model. He studied computer science and math at USC and Stanford, spent a few years at YouTube, and completed a PhD at UT Austin under Alex Dimakis. A 2021 paper on composing normalizing flows for inverse problems, written with Erik Lindgren and Dimakis, won a Best Paper Award at the UAI 2021 Workshop on Tractable Probabilistic Modeling. In 2022, the Imagen paper received a NeurIPS Outstanding Paper award, and he coauthored Imagen Video, a cascade of video diffusion models. He was among the authors of the Gemini 1.5 report in 2024.

Timeline

  1. By 2019

    USC, Stanford and YouTube

    He studies computer science and math at USC and Stanford and spends a few years at YouTube training classifiers and building backend infrastructure. In 2019, he coauthors a paper with Rui Shu, Hung Bui and Stefano Ermon on training variational autoencoders with Buffered Stochastic Variational Inference.

  2. 2021

    Normalizing flows for inverse problems

    With Erik Lindgren and Alex Dimakis, his PhD advisor at UT Austin, he publishes "Composing Normalizing Flows for Inverse Problems" at ICML in 2021. The paper wins a Best Paper Award at the UAI 2021 Workshop on Tractable Probabilistic Modeling.

  3. 2022

    Deblurring via stochastic refinement

    In 2022, he is first author of the CVPR paper "Deblurring via Stochastic Refinement", written with Mauricio Delbracio, Hossein Talebi, Chitwan Saharia, Alexandros G. Dimakis and Peyman Milanfar. A conditional diffusion model refines the output of a deterministic predictor, and the work is presented as an oral.

  4. 2022

    Core contributor to Imagen

    In 2022, he is a core contributor to Imagen, a text-to-image diffusion model from the Google Research Brain Team. The paper finds that generic language models such as T5, pretrained on text-only corpora, encode text well for image synthesis, and Imagen reaches a state-of-the-art FID score of 7.27 on the COCO dataset.

  5. 2022

    A NeurIPS Outstanding Paper

    The Imagen paper, "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding", receives an Outstanding Paper award at NeurIPS 2022. He is one of its fourteen authors.

  6. 2022

    Imagen Video

    In 2022, he coauthors Imagen Video, a text-conditional video generation system built on a cascade of video diffusion models that turns a text prompt into high definition video. He shares equal-contribution credit with Jonathan Ho, William Chan, Chitwan Saharia and Tim Salimans.

  7. 2024

    An author of Gemini 1.5

    In 2024, he is one of the authors of the Gemini 1.5 report from the Gemini Team at Google. The report introduces Gemini 1.5 Pro and Gemini 1.5 Flash, which reach near-perfect retrieval up to at least 10M tokens of context.

  8. As of October 2026

    Gemini at Google DeepMind

    He is a research scientist at Google DeepMind working on Gemini. In 2026, he joins a Google DeepMind panel at AGI House's Enterprise Build Day, where he makes the case for world models as the substrate for the next jump.

Career

  1. NowGoogle DeepMindResearch Scientist

Education

  1. University of Southern California
  2. Stanford University
  3. University of Texas at AustinPhD

From the biography.

  1. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Jay Whang. 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. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

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 Jay Whang yet. Send us a free-licensed photo.

  1. Jay Whang jaywhang.com. Checked 21 October 2026.
  2. Training Variational Autoencoders with Buffered Stochastic Variational Inference proceedings.mlr.press. Checked 21 October 2026.
  3. Composing Normalizing Flows for Inverse Problems proceedings.mlr.press. Checked 21 October 2026.
  4. Deblurring via Stochastic Refinement (CVPR 2022 Open Access Repository) openaccess.thecvf.com. Checked 21 October 2026.
  5. Imagen: Text-to-Image Diffusion Models imagen.research.google. Checked 21 October 2026.
  6. [2205.11487] Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding arxiv.org. Checked 21 October 2026.
  7. NeurIPS 2022 Awards neurips.cc. Checked 21 October 2026.
  8. [2210.02303] Imagen Video: High Definition Video Generation with Diffusion Models arxiv.org. Checked 21 October 2026.
  9. Imagen Video imagen.research.google. Checked 21 October 2026.
  10. [2403.05530] Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context arxiv.org. Checked 21 October 2026.
  11. 11 Days After I/O: Inside the Google DeepMind Enterprise Build Day blog.agihouse.org. Checked 21 October 2026.