Michal Lewandowski PhD

Senior Researcher Data Science

Publikationen

2026
Exploiting Space Folding by Neural Networks

Lewandowski, M., Pisoni, R., Heinzl, B., & Moser, B. A.. (2026). Exploiting Space Folding by Neural Networks. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 22833–22840). Association for the Advancement of Artificial Intelligence (AAAI). https://doi.org/10.1609/aaai.v40i27.39446
Zur Publikation
2025
On Tessellations of ReLU Neural Networks

Lewandowski, Michal. (2025). On Tessellations of ReLU Neural Networks. On Tessellations of ReLU Neural Networks
Zur Publikation
On Space Folds of ReLU Neural Networks

Lewandowski, M., Eghbalzadeh, H., Heinzl, B., Pisoni, R., & Moser, B. A. (2025). On Space Folds of ReLU Neural Networks. Transactions on Machine Learning Research.
2024
The Turing Game

Lewandowski, M., Schmid, S. L., Mederitsch, P., Aufreiter, A., Aichinger, G., Nessler, F., Bergsmann, S., Szolga, V., Halmdienst, T., & Nessler, B. (2024). The Turing Game. In The First Workshop on System-2 Reasoning at Scale, NeurIPS'24. Retrieved from https://openreview.net/forum?id=YiquAWqFw3
CantorNet - A Sandbox for Testing Geometrical and Topological Measures

Michal Lewandowski, Hamid Eghbalzadeh, & Bernhard A. Moser (2024). CantorNet: A Sandbox for Testing Topological and Geometrical Measures. In NeurIPS 2024 Workshop on Symmetry and Geometry in Neural Representations.
2022
Tessellation-Filtering ReLU Neural Networks

Moser, B. A., Lewandowski, M., Kargaran, S., Zellinger, W., Biggio, B., & Koutschan, C. (2022). Tessellation-Filtering ReLU Neural Networks. Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence.
Zur Publikation
2021
Discrimination properties of binarized relu activations

Lewandowski, M., Zellinger, W., Eghbal-zadeh, H., Shepeleva, N., & Moser, B. A. (2023). On the Discriminability of Samples Using Binarized ReLU Activations. ICDSMLA 2021, 65–71.
Zur Publikation
2020
Adversarial robustness in data augmentation

Eghbal-zadeh, H., Koutini, K., Haunschmid, V., Primus, P., Lewandowski, M., Zellinger, W., & Widmer, G. (2020) ICLR 2020 Workshop Towards Trustworthy ML: Rethinking Security and Privacy for ML.
Zur Publikation
ReLU code space: A basis for rating network quality besides accuracy

Shepeleva, Natalia, Zellinger, Werner, Lewandowski, Michal, & Moser, Bernhard. (2020). ReLU Code Space: A Basis for Rating Network Quality Besides Accuracy (Version 1). arXiv.
Zur Publikation