PeopleofSuperintelligence

Researcher

Stefanie Jegelka

Stefanie Jegelka is a machine-learning researcher whose work links combinatorial optimization with graph neural networks. She studied bioinformatics at Tübingen, completed a computer-science doctorate through Tübingen and ETH Zurich, and moved from a Berkeley postdoc to MIT's faculty. Her doctoral work modeled high-order graph interactions with submodular functions for machine-learning and computer-vision problems. After chairing ICML's 2022 program, she moved to TUM as a Humboldt Professor; at MCML, she is Core PI for Foundations of Deep Neural Networks and an ELLIS Fellow in Unit Munich.

Works
206
Citations
6,688
h-index
34

Publication numbers from OpenAlex, as of 14 October 2026.

Photo of Stefanie Jegelka

Timeline

  1. 2000s

    A first project in machine learning

    She studied bioinformatics at the University of Tübingen and earned her diploma in 2007. As an undergraduate, she worked on independent component analysis at the Max Planck Institute.

  2. 2012

    Modeling high-order interactions

    She completed a computer-science doctorate in 2012 through the University of Tübingen and ETH Zurich. Her thesis studied machine-learning and computer-vision problems with high-order graph interactions modeled as submodular functions over graph edges.

  3. 2012–2014

    Research in Berkeley

    From 2012 to 2014, she was a postdoctoral researcher in UC Berkeley's Department of Electrical Engineering and Computer Science.

  4. 2015

    Joining MIT's faculty

    After completing her Berkeley postdoc, she joined MIT's faculty as an assistant professor in 2015.

  5. 2022

    Leading ICML's program

    She served as a program chair for ICML in 2022.

  6. 2024

    Moving graph learning to Munich

    She and Suvrit Sra moved from MIT to TUM as Humboldt Professors in artificial intelligence. Her research includes optimization of graph neural networks.

  7. As of October 2026

    A research group in Munich

    At MCML, she is a Core PI for Foundations of Deep Neural Networks and a Humboldt Professor at TU Munich. Her research spans algorithmic machine learning, optimization, theory, and applications. She is an ELLIS Fellow in Unit Munich.

Sources and credits

Photo: mcml.ai; MCML - Research Group Stefanie Jegelka; creator not stated. License: Not stated by source. Source: mcml.ai. Cropped to a square and resized. Ask us to remove or replace this photo.

  1. Stefanie Jegelka | MIT CSAIL csail.mit.edu. Checked 14 October 2026.
  2. Stefanie Jegelka humboldt-foundation.de. Checked 14 October 2026.
  3. Combinatorial Problems with Submodular Coupling in Machine Learning and Computer Vision las.inf.ethz.ch. Checked 14 October 2026.
  4. Unpacking the “black box” to build better AI models news.mit.edu. Checked 14 October 2026.
  5. Stefanie Jegelka people.csail.mit.edu. Checked 14 October 2026.
  6. Two new Humboldt Professorships in the field of Artificial Intelligence tum.de. Checked 14 October 2026.
  7. Stefanie Jegelka humboldt-foundation.de. Checked 14 October 2026.
  8. MCML - Research Group Stefanie Jegelka mcml.ai. Checked 14 October 2026.
  9. Stefanie Jegelka | European Laboratory for Learning and Intelligent Systems ellis.eu. Checked 14 October 2026.