PeopleofSuperintelligence

Researcher

Tom Gunter

Tom Gunter is a machine learning researcher whose work spans Bayesian inference and large language models. He coauthored two 2014 papers on Cox-process modeling and probabilistic sampling, then completed an Oxford PhD thesis on Bayesian inference in 2017. At Apple, he was first author of the 2024 Apple Intelligence foundation-model report and left in June 2025 after eight years. He coauthored Apple's 2025 tech report and the 2026 Parallel Track Transformers paper, with the latter's work done while he was at Apple. Since July 2025, he is a researcher at Meta's Superintelligence Labs.

Badges

Timeline

  1. 2014

    Structured point processes

    In 2014, he coauthors Efficient Bayesian Nonparametric Modelling of Structured Point Processes, a UAI paper that presents a Bayesian generative model for dependent Cox point processes.

  2. 2014

    Fast Bayesian quadrature

    Also in 2014, he coauthors Sampling for Inference in Probabilistic Models with Fast Bayesian Quadrature, a NIPS paper that proposes an active-learning sampler for probabilistic inference with faster wall-clock convergence than Markov chain Monte Carlo.

  3. 2017

    An Oxford PhD thesis

    In 2017, he completes an Oxford PhD thesis on Bayesian inference, Cox processes and probabilistic integration.

  4. 2024

    Apple Intelligence models

    In July 2024, he is first author of Apple Intelligence Foundation Language Models, which describes a roughly 3-billion-parameter on-device model and a large server model for Private Cloud Compute.

  5. June 2025

    Leaving Apple

    In June 2025, he leaves Apple after eight years as one of its most senior large language model researchers.

  6. July 2025

    Apple's 2025 model report

    In July 2025, he coauthors the Apple Intelligence tech report, which describes a server model built on a novel Parallel-Track Mixture-of-Experts transformer.

  7. February 2026

    Parallel Track Transformers

    In February 2026, he coauthors Parallel Track Transformers, which reports up to a 16x reduction in synchronization operations relative to standard tensor parallelism. He does the work while at Apple.

  8. Since July 2025

    Joining Meta Superintelligence Labs

    Since July 2025, he works as an AI researcher in Meta's Superintelligence Labs after briefly working at another AI company. At Apple, he works closely with Ruoming Pang, the former head of Apple's Foundation Models team, who also joins Meta.

Career

  1. Since 2025Meta Superintelligence LabsResearcher
  2. Until 2025AppleLarge Language Model Researcher

Education

  1. 2017University of OxfordPhD

From the biography.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Tom Gunter. 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.
  2. 16% of AI career moves in the 2020s went from a university to a companyIncludes Tom Gunter. 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.
  3. 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.
  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 Tom Gunter yet. Send us a free-licensed photo.

  1. Efficient Bayesian Nonparametric Modelling of Structured Point Processes arxiv.org. Checked 21 October 2026.
  2. Sampling for Inference in Probabilistic Models with Fast Bayesian Quadrature arxiv.org. Checked 21 October 2026.
  3. Theses robots.ox.ac.uk. Checked 21 October 2026.
  4. Apple Intelligence Foundation Language Models arxiv.org. Checked 21 October 2026.
  5. Tom Gunter, one of Apple’s senior large language model researchers, has left the tech giant - Apple World Today appleworld.today. Checked 21 October 2026.
  6. Apple Intelligence Foundation Language Models: Tech Report 2025 arxiv.org. Checked 21 October 2026.
  7. Parallel Track Transformers: Enabling Fast GPU Inference with Reduced Synchronization arxiv.org. Checked 21 October 2026.
  8. Parallel Track Transformers: Enabling Fast GPU Inference with Reduced Synchronization - Apple Machine Learning Research machinelearning.apple.com. Checked 21 October 2026.
  9. Meta hires two Apple AI researchers for Superintelligence push, Bloomberg News reports finance.yahoo.com. Checked 21 October 2026.
  10. Meta is luring more top AI researchers from Apple with million-dollar deals the-decoder.com. Checked 21 October 2026.
  11. Apple loses fourth AI researcher in a month to Meta’s superintelligence team latimes.com. Checked 21 October 2026.
  12. Meta just poached two more AI researchers from Apple amid internal shakeup 9to5mac.com. Checked 21 October 2026.