PeopleofSuperintelligence

Researcher

Rachel Hornung

Rachel Hornung is a software engineer and researcher at Google DeepMind whose work centers on large multimodal and video models. They studied business information technology in Mannheim and robotics at the Technical University of Munich, then spent 2012 to 2018 as a researcher at the German Aerospace Center (DLR), where they coauthored a 2014 paper on robot anomaly detection. After joining Google in 2019, they coauthored the 2023 NeurIPS paper on Integrated Multimodal Perception and VideoPoet, a language model for video generation that received a Best Paper award at ICML 2024. They also coauthored the VideoPrism video encoder, the Neptune long-video benchmark and the Gemini 2.5 report, and since 2024 they are a Staff Software Engineer at Google DeepMind.

Timeline

  1. 2007–2012

    Degrees in Mannheim and Munich

    They study International Business Information Technology at Duale Hochschule Baden-Württemberg Mannheim from 2007 to 2010, then Robotics, Cognition, Intelligence at the Technical University of Munich from 2010 to 2012. Their 2012 master's thesis, "Single-Class Discrimination for Robot Anomaly Detection", introduces a Robot Anomaly Detection System that finds errors without previous records of them.

  2. 2012–2018

    Robot anomaly detection at DLR

    They work as a researcher at the German Aerospace Center (DLR) from 2012 to 2018. In 2014, they coauthor "Model-free robot anomaly detection" with Holger Urbanek, Julian Klodmann, Christian Osendorfer and Patrick van der Smagt for the IROS conference in Chicago; the paper introduces a robot anomaly detection system for data with little or no anomaly information.

  3. 2019

    Joining Google

    In 2019, they join Google, and their chapter "Early integration for movement modeling in latent spaces" appears in The Handbook of Multimodal-Multisensor Interfaces.

  4. 2023

    Integrated Multimodal Perception at NeurIPS

    In 2023, they coauthor "Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception" with Hassan Akbari, Dan Kondratyuk, Yin Cui, Huisheng Wang and Hartwig Adam, published at NeurIPS 2023. The IMP model feeds image, video, text and audio into a single Transformer encoder, and its sparse IMP-MoE-L variant sets a new state of the art in zero-shot video classification with 77.0% on Kinetics-400.

  5. December 2023

    VideoPoet, a language model for video

    In December 2023, they are among the coauthors of VideoPoet, a Google Research language model that synthesizes high-quality video with matching audio from a large variety of conditioning signals. It animates input images, edits video by inpainting and outpainting, and stylizes video.

  6. 2024

    Best Paper at ICML 2024

    VideoPoet receives a Best Paper award at ICML 2024, with Hornung among its listed authors.

  7. 2024

    VideoPrism and the Neptune benchmark

    In February 2024, they appear among the authors of VideoPrism, a general-purpose video encoder that reaches state-of-the-art performance on 31 of 33 video understanding benchmarks with a single frozen model. In December 2024, they coauthor Neptune, a benchmark for long video understanding built on video segments of up to 15 minutes.

  8. Since 2024

    Staff engineer at Google DeepMind

    In 2024, they take the role of Staff Software Engineer at Google DeepMind. In 2025, they are among the listed authors of the Gemini 2.5 report, which introduces Gemini 2.5 Pro and Gemini 2.5 Flash.

Career

  1. Since 2024Google DeepMindStaff Software Engineer
  2. From 2019Google
  3. 2012–2018German Aerospace CenterResearcher

Education

  1. 2012Technical University of MunichMaster's degree
  2. 2010Baden-Württemberg Cooperative State University Mannheim

From the biography.

  1. 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.
  2. 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.
  3. 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.
  4. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Google. 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.

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 Rachel Hornung yet. Send us a free-licensed photo.

  1. Rachel Hornung (Google) — Lacuna lacuna.tiptreesystems.com. Checked 22 October 2026.
  2. electronic library - Single-Class Discrimination for Robot Anomaly Detection elib.dlr.de. Checked 22 October 2026.
  3. electronic library - Model-free robot anomaly detection elib.dlr.de. Checked 22 October 2026.
  4. Rachel Hornung - Google Scholar scholar.google.com. Checked 22 October 2026.
  5. Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception proceedings.neurips.cc. Checked 22 October 2026.
  6. [2312.14125] VideoPoet: A Large Language Model for Zero-Shot Video Generation arxiv.org. Checked 22 October 2026.
  7. VideoPoet: A large language model for zero-shot video generation research.google. Checked 22 October 2026.
  8. ICML 2024 Awards icml.cc. Checked 22 October 2026.
  9. [2402.13217] VideoPrism: A Foundational Visual Encoder for Video Understanding arxiv.org. Checked 22 October 2026.
  10. [2412.09582] Neptune: The Long Orbit to Benchmarking Long Video Understanding arxiv.org. Checked 22 October 2026.
  11. [2507.06261] Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities arxiv.org. Checked 22 October 2026.