PeopleofSuperintelligence

Researcher

Daniel Zügner

Daniel Zügner is a Senior Researcher at Microsoft Research AI for Science whose work spans graph learning and materials discovery. He coauthored early studies of adversarial attacks on graph neural networks and completed a 2022 PhD at the Technical University of Munich. At Microsoft, he helped build MatterGen's machine-learning architecture for property-guided materials design; its Nature paper reported an experimentally synthesized structure with a measured property value within 20% of its target. In May 2026, he coauthored MatterSim-MT, a model for materials simulation and property characterization trained on more than 35 million structures.

Now at
Microsoft
Works
36
Citations
3,223
h-index
16
Badges

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

  1. 2018

    Attacking graph classifiers

    In 2018, he and Amir Akbarnejad and Stephan Günnemann study adversarial attacks on attributed graphs and introduce Nettack, an algorithm for perturbing node features and graph structure.

  2. 2019

    Optimizing graph attacks

    In 2019, he and Stephan Günnemann use meta-gradients to optimize training-time attacks that perturb graph structure in graph neural networks.

  3. 2022

    A doctorate in graph learning

    In 2022, he completes a PhD at the Technical University of Munich, specializing in machine learning and graph neural networks, and begins his work in AI for Science.

  4. December 2023

    Designing materials with MatterGen

    In December 2023, he coauthors a Microsoft Research paper introducing MatterGen, a diffusion model that generates stable inorganic materials and can be fine-tuned for target-property constraints.

  5. January 2025

    MatterGen's Nature paper

    In January 2025, he contributes to MatterGen's Nature paper by helping build its machine-learning architecture for property-guided design. The team reports an experimentally synthesized generated structure with a measured property value within 20% of its target.

  6. May 2026

    Extending materials simulation

    In May 2026, he coauthors MatterSim-MT, a multi-task model for in silico materials simulation and property characterization trained on more than 35 million first-principles-labeled structures.

  7. As of October 2026

    Senior Researcher at Microsoft

    As of October 2026, he is a Senior Researcher at Microsoft Research AI for Science.

Career

  1. Since 2023MicrosoftSenior Researcher

Education

  1. 2022Technical University of MunichPhD

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Adversarial Attacks on Neural Networks for Graph Data2018 · ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (KDD) · 973 citationsWith Stephan Günnemann
  2. A generative model for inorganic materials design2025 · Nature · 508 citationsWith Andrew Fowler and Claudio Zeni
  3. Adversarial Attacks on Graph Neural Networks via Meta Learning2019 · 351 citationsWith Stephan Günnemann
  4. NetGAN: Generating Graphs via Random Walks2018 · 251 citationsWith Stephan Günnemann
  5. Adversarial Attacks on Graph Neural Networks2020 · ACM Transactions on Knowledge Discovery from Data · 117 citationsWith Stephan Günnemann

From OpenAlex, as of 19 October 2026.

  1. 14% of AI career moves in the 2020s went from a university to a companyIncludes Daniel Zügner. Our estimates of how university-to-company career transitions changed by decade, including roles added alongside an earlier job.
  2. The company share of AI researchers’ top papers rose from 2% to 32%Includes Microsoft. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 2% in 1975–1979 to 32% in 2020–2024.
  3. Google has the highest combined h-index among AI organizationsIncludes Microsoft. Our estimates of combined and median h-index at AI organizations, using current affiliations and OpenAlex records.
  4. Google alumni founded or led 87 other organizations, the most among the labs we compareIncludes Microsoft. We trace 212 organizations founded or led by AI lab alumni, and 223 career moves between the labs.

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 Daniel Zügner yet. Send us a free-licensed photo.

  1. Adversarial Attacks on Neural Networks for Graph Data arxiv.org. Checked 19 October 2026.
  2. Adversarial Attacks on Graph Neural Networks via Meta Learning arxiv.org. Checked 19 October 2026.
  3. AI meets materials discovery: The vision behind MatterGen and MatterSim microsoft.com. Checked 19 October 2026.
  4. MatterGen: a generative model for inorganic materials design microsoft.com. Checked 19 October 2026.
  5. A generative model for inorganic materials design nature.com. Checked 19 October 2026.
  6. Advancing AI for materials with MatterSim: experimental synthesis, faster simulation, and multi-task models microsoft.com. Checked 19 October 2026.
  7. MatterSim-MT: A multi-task foundation model for in silico materials characterization arxiv.org. Checked 19 October 2026.
  8. Daniel Zügner at Microsoft Research microsoft.com. Checked 19 October 2026.