Steven Braun

I received my PhD from the Artificial Intelligence and Machine Learning Lab, TU Darmstadt. My main research interests cover a broad range of Machine Learning related topics such as deep models, tractable probabilistic models such as probabilistic circuits, and their applications. In specific, I work on bridging the gap between probabilistic circuits and deep neural networks. We want to push the limits of probabilistic circuits and aim to combine their strenghts of tractable flexibility with the modeling capacity of neural networks.

GitHub · Google Scholar · ORCID · Mastodon · Twitter

Blog

• 2024-11-15 Emacs Org Mode: org-emphasize-dwim
• 2022-10-06 Tiling macOS: Moving from i3wm to Yabai
• 2021-10-27 Optimizing Matplotlib Visualizations for Academic Papers
• 2021-09-21 A Short Review of Axis-Aligned and Oriented Object Detection
• 2021-07-12 The arXiv PDF Command-Line Interface Downloader

Publications

• Bridging Probabilistic Circuits and Deep Neural Networks — Ph.D. Thesis, Technische Universität Darmstadt, 2026
• Tractable Representation Learning with Probabilistic Circuits — Transactions on Machine Learning Research (TMLR), 2025
• Deep Classifier Mimicry without Data Access — International Conference on Artificial Intelligence and Statistics (AISTATS) – Oral & Student Paper Highlight Award, 2024
• Probabilistic Circuits That Know What They Don’t Know — Proceedings of the 39th Conference on Uncertainty in Artificial Intelligence (UAI), 2023
• Towards Coreset Learning in Probabilistic Circuits — The 5th Workshop on Tractable Probabilistic Modeling (UAI), 2022
• CLEVA-Compass: A Continual Learning EValuation Assessment Compass to Promote Research Transparency and Comparability — International Conference on Learning Representations (ICLR), 2022
• Elevating Perceptual Sample Quality in Probabilistic Circuits through Differentiable Sampling — Proceedings of Machine Learning Research, Workshop on Preregistration in Machine Learning (NeurIPS), 2022
• DAFNe: A One-Stage Anchor-Free Deep Model for Oriented Object Detection — arXiv preprint, arXiv:2109.06148, 2021
• Einsum Networks: Fast and Scalable Learning of Tractable Probabilistic Circuits — Proceedings of the 37th International Conference on Machine Learning (ICML), 2020