PeopleofSuperintelligence

Researcher

Pei Sun

Pei Sun is an AI researcher whose work connects autonomous-driving perception with multimodal and reasoning models. He was first author of the Waymo Open Dataset paper, coauthored range-image and sparse-window 3D detectors, and supported Waymo’s automated data-augmentation work. At Waymo, he was a Research Scientist scheduled to speak about scaling 3D detection toward end-to-end driving; he later coauthored Gemini 1.5 and contributed to post-training for the 2025 Deep Think IMO effort. Since 2025, he is part of Meta Superintelligence Labs, after working on Gemini post-training, coding, and reasoning at Google DeepMind.

Works
20

Publication numbers from OpenAlex, as of 22 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2019

    Waymo Open Dataset paper

    In 2019, he is first author of “Scalability in Perception for Autonomous Driving: Waymo Open Dataset,” which introduces 1,150 20-second scenes of synchronized LiDAR and camera data.

  2. 2020

    Automated data augmentation

    In 2020, he supports Waymo and Google Research’s automated data-augmentation work on perception for the Waymo Driver.

  3. 2021

    To the Point detector

    In 2021, he coauthors “To the Point,” which learns a 3D representation from range images and improves pedestrian average precision on the Waymo Open Dataset from 69.7% to 75.5%.

  4. 2022

    SWFormer 3D detector

    In 2022, he coauthors SWFormer, which uses sparse voxels and windows for 3D detection and reaches 73.36 L2 mAPH for vehicles and pedestrians on the Waymo Open Dataset official test set.

  5. 2023

    Waymo Research Scientist

    In 2023, he is Waymo’s Research Scientist scheduled to speak at the End-to-End Autonomous Driving Workshop about scaling 3D detection toward end-to-end driving.

  6. March 2024

    Gemini 1.5 report

    In March 2024, he coauthors the Gemini 1.5 technical report, which introduces multimodal models that recall and reason over millions of tokens of context.

  7. 2025

    Gemini Deep Think post-training

    In 2025, he contributes to post-training for the advanced Gemini model with Deep Think that reaches a gold-medal score of 35 out of 42 points at the International Mathematical Olympiad.

  8. Since 2025

    Meta Superintelligence Labs

    In 2025, he joins Meta Superintelligence Labs. At Google DeepMind, he works on Gemini post-training, coding, and reasoning after creating the last two generations of Waymo’s perception models.

Career

  1. Since 2025Meta Superintelligence Labs
  2. WaymoResearch Scientist

From the biography.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Pei Sun. 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.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Meta. 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 Meta Superintelligence Labs. 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. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes Meta Superintelligence Labs. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.

Colleagues

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

Sources and credits

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

  1. Scalability in Perception for Autonomous Driving: Waymo Open Dataset arxiv.org. Checked 21 October 2026.
  2. Using automated data augmentation to advance our Waymo Driver waymo.com. Checked 21 October 2026.
  3. To the Point: Efficient 3D Object Detection in the Range Image with Graph Convolution Kernels arxiv.org. Checked 21 October 2026.
  4. SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds arxiv.org. Checked 21 October 2026.
  5. Waymo at CVPR 2023: At the cutting-edge of autonomous driving research waymo.com. Checked 21 October 2026.
  6. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context arxiv.org. Checked 21 October 2026.
  7. Advanced version of Gemini with Deep Think officially achieves gold-medal standard at the International Mathematical Olympiad deepmind.google. Checked 21 October 2026.
  8. Here Is Everyone Mark Zuckerberg Has Hired So Far for Meta’s ‘Superintelligence’ Team wired.com. Checked 21 October 2026.
  9. Meta’s new hires offer a peek into superintelligence plans semafor.com. Checked 21 October 2026.