Researcher

Helmut Bölcskei

Helmut Bölcskei is a professor at ETH Zurich whose research connects mathematical information science with machine learning theory. He earned electrical engineering degrees at Vienna University of Technology, worked in research at Stanford and held an Assistant Professor post at Illinois, helped found a MIMO wireless startup, and joined ETH Zurich in 2002. His work spans interference-channel limits and neural-network identifiability, including a 2020 paper with Verner Vlačić on affine symmetries and neural-network identifiability. He received a Rothschild Fellowship in 2021 and gave a Munich AI Lecture in 2024; as of October 2026, he is a Professor of Mathematical Information Science at ETH Zurich.

Timeline

  1. 1994–1997

    Engineering degrees in Vienna

    He earns a Dipl.-Ing. degree in electrical engineering from Vienna University of Technology in 1994 and a Dr. techn. degree in 1997.

  2. 1999–2002

    Stanford and Illinois posts

    He works as a postdoctoral researcher in Stanford University's Information Systems Laboratory and Department of Statistics from 1999 to 2001. From 2001 to 2002, he serves as an Assistant Professor of Electrical Engineering at the University of Illinois at Urbana-Champaign.

  3. 2002

    Intel acquires Iospan Wireless

    Intel acquires Iospan Wireless in 2002. He was part of the founding team of the Silicon Valley startup, which focused on MIMO systems for high-speed Internet access.

  4. 2014

    Interference alignment conditions

    At a Stanford colloquium in 2014, he presents conditions under which K-user single-antenna interference channels achieve K/2 degrees of freedom for almost all channel matrices. He connects interference alignment with additive combinatorics.

  5. 2020

    Neural network identifiability

    With Verner Vlačić, he studies neural-network identifiability in a 2020 paper. They use affine symmetries to study networks that produce the same function and derive identifiability results for a class of tanh-type nonlinearities.

  6. 2021–2024

    A fellowship and AI lecture

    He receives a Rothschild Fellowship from the Isaac Newton Institute for Mathematical Sciences at Cambridge in 2021 and gives a Munich AI Lecture in 2024.

  7. As of October 2026

    Professor of Mathematical Information Science

    As of October 2026, he is Professor of Mathematical Information Science at ETH Zurich; he has been with the university since 2002 and is associated with its Departments of Information Technology and Electrical Engineering, Mathematics, and Computer Science. His research interests include machine learning theory, applied mathematics, mathematical signal processing, data science, and statistics.

Career

  1. Since 2002ETH ZurichProfessor of Mathematical Information Science, Core Faculty, ETH AI Center
  2. 2021–2026Lagrange Mathematics and Computing Research CenterPrincipal Investigator
  3. 2001–2002University of Illinois Urbana-ChampaignAssistant Professor of Electrical Engineering
  4. 1999–2001Stanford UniversityPostdoctoral Researcher
  5. Iospan WirelessFounding Team Member
  6. Celestrius AGCo-founder
  7. Philips Research LaboratoriesVisiting Researcher
  8. Télécom ParisVisiting Researcher
  9. Heinrich Hertz InstituteVisiting Researcher

Education

  1. 1997TU WienDr. techn.
  2. 1994TU WienDipl.-Ing.

Sources and credits

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

  1. Chair for Mathematical Information Science > Home mins.ee.ethz.ch. Checked 26 October 2026.
  2. Explicit and almost sure conditions for K/2 degrees of freedom isl.stanford.edu. Checked 26 October 2026.
  3. Affine symmetries and neural network identifiability arxiv.org. Checked 26 October 2026.
  4. Faculty – ETH AI Center | ETH Zurich ai.ethz.ch. Checked 26 October 2026.
  5. Chair for Mathematical Information Science > People mins.ee.ethz.ch. Checked 26 October 2026.