Current Group Member

Current and Upcoming Projects

Multilevel Architectures and Algorithms in Deep Learning
Start: 2023-01-01
End: 2025-12-31
Principal Investigators: Roland Herzog, Anton Schiela
Staff: Leonie Kreis, Frederik Köhne
Funded by: DFG within the Priority Program funding scheme
Part of: Theoretical Foundations of Deep Learning (SPP 2298)

Recent Completed Projects

Latest Publications

  • Roland Herzog, Frederik Köhne, Leonie Kreis and Anton Schiela
    Metric Frobenius norms and inner products of matrices and linear maps
    Linear Algebra and its Applications, 2025
    bibtex
    @ARTICLE{HerzogKoehneKreisSchiela:2025:1,
      AUTHOR = {Herzog, Roland and Köhne, Frederik and Kreis, Leonie and Schiela, Anton},
      DATE = {2025-08-08},
      DOI = {10.1016/j.laa.2025.08.005},
      EPRINT = {2311.15419},
      EPRINTTYPE = {arXiv},
      JOURNALTITLE = {Linear Algebra and its Applications},
      TITLE = {Metric Frobenius norms and inner products of matrices and linear maps},
    }
  • Evelyn Herberg, Roland Herzog, Frederik Köhne, Leonie Kreis and Anton Schiela
    Sensitivity-based layer insertion for residual and feedforward neural networks, 2023
    bibtex
    @ONLINE{HerbergHerzogKoehneKreisSchiela:2023:1,
      AUTHOR = {Herberg, Evelyn and Herzog, Roland and Köhne, Frederik and Kreis, Leonie and Schiela, Anton},
      DATE = {2023-11},
      EPRINT = {2311.15995},
      EPRINTTYPE = {arXiv},
      TITLE = {Sensitivity-based layer insertion for residual and feedforward neural networks},
    }
  • Frederik Köhne, Leonie Kreis, Anton Schiela and Roland Herzog
    Adaptive step sizes for preconditioned stochastic gradient descent, 2023
    bibtex
    @ONLINE{KoehneKreisSchielaHerzog:2023:1,
      AUTHOR = {Köhne, Frederik and Kreis, Leonie and Schiela, Anton and Herzog, Roland},
      DATE = {2023-11},
      EPRINT = {2311.16956},
      EPRINTTYPE = {arXiv},
      TITLE = {Adaptive step sizes for preconditioned stochastic gradient descent},
    }

Recent Teaching