Nils Philipp Walter

I am a second year Ph.D. student at CISPA Helmholtz Center for Information Security, supervised by Jilles Vreeken. I am broadly interested in robust and explainable machine learning for large-scale real-world applications. In my Ph.D, I intend to develop new approaches that are at the same time descriptive and predictive. That is the models not only offer predictive capabilities but also facilitate practitioners to gain deeper insights into the problems they are addressing. Currently, I mostly work on understanding how neural networks process information and building methods to describe when and how models make errors.

Before joining CISPA, I was a research assistant in the goup of Bernt Schiele at the Max-Planck-Institut for Informatics, supervised by David Stutz . My research focused on adversarial and out-of-distribution robustness of Quantized Neural Networks. I also worked on the influence of Batch Normalization on the vulnerability and generalization capabilities of neural networks.

news

Mar '25	Preprint of our paper Now you see me! A framework for obtaining class-relevant saliency maps is available on arXiv.
Nov '24	I was invited to give a talk at the Efficient Machine Learning Reading Group . The slides are available here and a recording can be found on YouTube .
Nov '24	A new preprint of our work on learning Rule list classifiers in a fully differntiable manner is available on arXiv .
Sep '24	I gave a talk at the Gutenberg Workshop on AI for Scientific Discovery. The slides are available here .
Jun '24	Happy to announce that our paper Learning Exceptional Subgroups by End-to-End Maximizing KL-divergence received a spotlight at ICML 24! 🎉
May '24	Preprint of our paper The Uncanny Valley: Exploring Adversarial Robustness from a Flatness Perspective is available on arXiv.
May '24	Our paper Learning Exceptional Subgroups by End-to-End Maximizing KL-divergence got accepted to ICML 24!
Feb '24	Preprint of our paper Learning Exceptional Subgroups by End-to-End Maximizing KL-divergence is available on arXiv.
Feb '24	Attending AAAI 2024 in Vancouver, Canada.
Dec '23	Attended Kickoff-Retreat of the RTG for Neuroexplicit Models at Schloss Dagstuhl.

publications

2025

arXiv

Now you see me! A framework for obtaining class-relevant saliency maps

Nils Philipp Walter, Jilles Vreeken, and Jonas Fischer

arXiv preprint arXiv:2503.07346, 2025

Bib HTML PDF

@article{walter2025var,
  title = {Now you see me! A framework for obtaining class-relevant saliency maps},
  author = {Walter, Nils Philipp and Vreeken, Jilles and Fischer, Jonas},
  journal = {arXiv preprint arXiv:2503.07346},
  year = {2025},
  organization = {arXiv},
}

2024

arXiv

Neuro-Symbolic Rule Lists

Sascha Xu, Nils Philipp Walter, and Jilles Vreeken

arXiv preprint arXiv:2411.06428, 2024

Bib HTML PDF

@article{walter2024neurules,
  title = {Neuro-Symbolic Rule Lists},
  author = {Xu, Sascha and Walter, Nils Philipp and Vreeken, Jilles},
  journal = {arXiv preprint  arXiv:2411.06428},
  year = {2024},
  organization = {arXiv},
}

arXiv

The Uncanny Valley: Exploring Adversarial Robustness from a Flatness Perspective

Nils Philipp Walter, Linara Adilova, Jilles Vreeken, and Michael Kamp

arXiv preprint arXiv:2405.16918, 2024

Bib HTML PDF

@article{walter2024uncanny,
  title = {The Uncanny Valley: Exploring Adversarial Robustness from a Flatness Perspective},
  author = {Walter, Nils Philipp and Adilova, Linara and Vreeken, Jilles and Kamp, Michael},
  journal = {arXiv preprint  arXiv:2405.16918},
  year = {2024},
  organization = {arXiv},
}

ICML

Learning Exceptional Subgroups by End-to-End Maximizing KL-divergence

Sascha Xu, Nils Philipp Walter, Janis Kalofolias, and Jilles Vreeken

In Proceedings of the International Conference on Machine Learning (ICML), 2024

Bib PDF

@inproceedings{xu2024learningicml,
  title = {Learning Exceptional Subgroups by End-to-End Maximizing KL-divergence},
  author = {Xu, Sascha and Walter, Nils Philipp and Kalofolias, Janis and Vreeken, Jilles},
  booktitle = {Proceedings of the International Conference on Machine Learning (ICML)},
  year = {2024},
  organization = {PMLR},
}

AAAI

Finding Interpretable Class-Specific Patterns through Efficient Neural Search

Nils Philipp Walter, Jonas Fischer, and Jilles Vreeken

In Proceedings of the 38th Annual AAAI Conference on Artificial Intelligence, 2024

Bib HTML PDF

@inproceedings{walter2024finding,
  title = {Finding Interpretable Class-Specific Patterns through Efficient Neural Search},
  author = {Walter, Nils Philipp and Fischer, Jonas and Vreeken, Jilles},
  booktitle = {Proceedings of the 38th Annual AAAI Conference on Artificial Intelligence},
  year = {2024},
  organization = {AAAI},
}

2022

CVPR

On Fragile Features and Batch Normalization in Adversarial Training

Nils Philipp Walter, David Stutz, and Bernt Schiele

arXiv preprint arXiv:2204.12393, 2022

Bib HTML PDF

@article{walter2022fragile,
  title = {On Fragile Features and Batch Normalization in Adversarial Training},
  author = {Walter, Nils Philipp and Stutz, David and Schiele, Bernt},
  journal = {arXiv preprint arXiv:2204.12393},
  year = {2022},
}