PeopleofSuperintelligence

Researcher

Mario Lucic

Also known as Mario Lučić

Mario Lucic is a machine learning researcher whose work spans generative models, representation learning, vision, and video understanding. Before Google, he completed a PhD at ETH Zurich on efficient data selection and clustering; by 2019, he worked at Google Research in Zurich. He coauthored a large-scale GAN study, a 2019 ICML Best Paper on unsupervised disentangled representations, MLP-Mixer, and PaLI-X. In May 2025, he was a Research Scientist and Gemini video-understanding lead; as of October 2026, he is a Research Director at Google DeepMind leading research teams across video understanding and agentic vision.

Timeline

  1. By 2019

    PhD at ETH Zurich

    He completes a PhD at ETH Zurich on efficient data selection and clustering before joining Google. By 2019, he works as a researcher at Google Research in Zurich.

  2. 2018–2019

    Testing GANs at scale

    In 2018, he coauthors A Large-Scale Study on Regularization and Normalization in GANs. The study examines training pitfalls and reproducibility issues, and its revised version is accepted to ICML 2019.

  3. 2019

    An ICML Best Paper

    In 2019, he coauthors Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations, which wins ICML's Best Paper Award. The study argues that unsupervised learning of disentangled representations is fundamentally impossible without inductive biases on the models and data.

  4. 2021

    MLP-Mixer for vision

    In 2021, he coauthors MLP-Mixer, a vision architecture built exclusively from multi-layer perceptrons. Its design applies MLPs independently to image patches and across patches, and NeurIPS presents the paper as a poster.

  5. 2024

    PaLI-X at CVPR

    In 2024, he coauthors PaLI-X, a multilingual vision-and-language model evaluated on image captioning, question answering, document understanding, object detection, and video tasks. The paper reports results across more than 20 vision-and-language benchmarks.

  6. May 2025

    Gemini 2.5 video work

    In May 2025, he serves as a Research Scientist and video-understanding lead for Gemini 2.5. Gemini 2.5 combines audio-visual information with code to support video applications.

  7. As of October 2026

    Research Director at Google DeepMind

    As of October 2026, he is a Research Director at Google DeepMind, leading research teams across video understanding, agentic vision, and compact encoders for Gemini. In September 2026, he coauthors Google's launch of agentic video understanding, which searches video frames, audio, and transcripts.

Career

  1. Since 2025Google DeepMindResearch Scientist and video-understanding lead
  2. From 2019Google ResearchResearcher

Education

  1. ETH ZurichPhD

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 Mario Lucic. 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 with dated positions at the same organization in the same years.

Sources and credits

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

  1. Mario Lučić | Research & Gemini Video Demos lucic.ai. Checked 21 October 2026.
  2. ICML Best Paper Award for Francesco Locatello ethz.ch. Checked 21 October 2026.
  3. A Large-Scale Study on Regularization and Normalization in GANs arxiv.org. Checked 21 October 2026.
  4. Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations research.google. Checked 21 October 2026.
  5. MLP-Mixer: An All-MLP Architecture for Vision research.google. Checked 21 October 2026.
  6. PaLI-X: On Scaling up a Multilingual Vision and Language Model research.google. Checked 21 October 2026.
  7. Advancing the frontier of video understanding with Gemini 2.5 developers.googleblog.com. Checked 21 October 2026.
  8. Introducing agentic video understanding with Gemini blog.google. Checked 21 October 2026.