Publications

Leena Chennuru Vankadara*, Moritz Haas*, Luke Hayward†, Sebastian Bordt†, Alessandro Breccia† (2026). How to Scale Mixture-of-Experts: From μP to the Maximally Scale-Stable Parameterization. Preprint; Oral at ICML HiLD Workshop 2026.

PDF Cite Code

Moritz Haas, Sebastian Bordt, Ulrike Von Luxburg, Leena Chennuru Vankadara (2025). On the Surprising Effectiveness of Large Learning Rates under Standard Width Scaling. NeurIPS 2025 (spotlight).

PDF Cite Code Blog

Leena Chennuru Vankadara*, Jin Xu*, Moritz Haas, Volkan Cevher (2024). On Feature Learning in Structured State Space Models. NeurIPS 2024. Oral and runner-up for best paper award at ICML NGSM Workshop 2024..

PDF Cite

Moritz Haas, Jin Xu, Volkan Cevher, Leena Chennuru Vankadara (2024). mup^2: Effective Sharpness Aware Minimization Requires Layerwise Perturbation Scaling. NeurIPS 2024.

PDF Cite

Moritz Haas*, David Holzmüller*, Ulrike Luxburg, Ingo Steinwart (2023). Mind the spikes: Benign overfitting of kernels and neural networks in fixed dimension. NeurIPS 2023.

PDF Cite Code Poster Slides

Moritz Haas, Bedartha Goswami, Ulrike Luxburg (2023). Pitfalls of Climate Network Construction - A Statistical Perspective. Journal of Climate.

PDF Cite Code Poster

Moritz Haas, Stefan Richter (2020). Statistical Analysis of Wasserstein GANs with Applications to Time Series Forecasting. arXiv:2011.03074.

PDF Cite Poster Slides