Publications

Understanding and Mitigating Exploding Inverses in Invertible Neural Networks

arXiv preprint
Jens Behrmann, Paul Vicol, Kuan-Chieh Wang, Roger Grosse, Joern-Henrik Jacobsen
PDF Code
Understanding and Mitigating Exploding Inverses in Invertible Neural Networks

Fundamental Tradeoffs between Invariance and Sensitivity to Adversarial Perturbations

In International Conference on Machine Learning (ICML)
Florian Tramer, Jens Behrmann, Nicholas Carlini, Nicolas Papernot, Joern-Henrik Jacobsen
PDF Code
Fundamental Tradeoffs between Invariance and Sensitivity to Adversarial Perturbations

Conditional Normalizing Flows for Low-Dose Computed Tomography Image Reconstruction

In ICML INNF+ workshop
Alexander Denker, Maximilian Schmidt, Johannes Leuschner, Peter Maass, Jens Behrmann
PDF
Conditional Normalizing Flows for Low-Dose Computed Tomography Image Reconstruction

Residual Flows for Invertible Generative Modeling

In Neural Information Processing Systems (NeurIPS)
Ricky T. Q. Chen, Jens Behrmann, David Duvenaud, Joern-Henrik Jacobsen
PDF Code
Residual Flows for Invertible Generative Modeling

Deep Relevance Regularization: Interpretable and Robust Tumor Typing of Imaging Mass Spectrometry Data

arXiv preprint
Christian Etmann, Maximilian Schmidt, Jens Behrmann, Tobias Boskamp, Lena Hauberg-Lotte, Annette Peter, Rita Casadonte, Joerg Kriegsmann, Peter Maass
PDF Project
Deep Relevance Regularization: Interpretable and Robust Tumor Typing of Imaging Mass Spectrometry Data

Invertible Residual Networks

In International Conference on Machine Learning (ICML)
Jens Behrmann, Will Grathwohl, Ricky T. Q. Chen, David Duvenaud, Joern-Henrik Jacobsen
PDF Code
Invertible Residual Networks

Excessive Invariance Causes Adversarial Vulnerability

In International Conference on Learning Represenations (ICLR)
Joern-Henrik Jacobsen, Jens Behrmann, Richard Zemel, Matthias Bethge
PDF
Excessive Invariance Causes Adversarial Vulnerability

Deep learning for tumor classification in imaging mass spectrometry

In Bioinformatics
Jens Behrmann, Christian Etmann, Tobias Boskamp, Rita Casadonte, Joerg Kriegsmann, Peter Maass
PDF Code Dataset Project
Deep learning for tumor classification in imaging mass spectrometry