Timo Heiß

About

I am a PhD student in the Machine Learning Interpretability subgroup under supervision of Giuseppe Casalicchio and Bernd Bischl since October 2025. My research interests center on making Machine Learning interpretable to increase trust and transparency, as well as on intersections of Interpretable Machine Learning with other areas of Machine Learning.

My current research focuses on feature effects, feature interactions, and functional decompositions. Previously, I have also worked on prompt optimization.

I obtained a Master's degree in "Statistics & Data Science" at LMU Munich and a Bachelor's degree in "Business Informatics - Data Science" at DHBW Ravensburg. Prior to my PhD studies, I gained five years of industry experience at Liebherr-Aerospace as a dual and working student, developing and deploying Machine Learning models with focus on interpretability.

Contact

Institut für Statistik

Ludwig-Maximilians-Universität München

Ludwigstraße 33

D-80539 München

Room 344, 3rd floor

timo.heiss [at] stat.uni-muenchen.de

Research Interests

• Interpretable Machine Learning
• Model-agnostic IML Methods
• Feature Effects, Interactions, and Functional Decompositions
• Intersections of IML
• (Prompt Optimization)

You can find me on

• Google Scholar
• GitHub
• LinkedIn

References

1. Heiß T, Bögel C, Bischl B, Casalicchio G (2026) Analyzing Error Sources in Global Feature Effect Estimation. arXiv:2603.15057 [stat.ML], Accepted at XAI 2026.
   arXiv | PDF | Code.
2. Zehle T, Heiß T, Schlager M, Aßenmacher M, Feurer M (2025) promptolution: A Unified, Modular Framework for Prompt Optimization. arXiv:2512.02840 [cs.CL], Accepted at EACL 2025.
   arXiv | PDF | GitHub | Documentation.
3. Zehle T, Schlager M, Heiß T, Feurer M (2025) CAPO: Cost-Aware Prompt Optimization. In: In: Akoglu L , In: Doerr C , In: Rijn JN van , In: Garnett R , In: Gardner JR (eds) Proceedings of the Fourth International Conference on Automated Machine Learning, pp. 18/1–45. PMLR.
   Link | PDF | arXiv | Code.