Florian P.
Mahner

PhD Candidate in Machine Learning

I study representation learning across humans and machines—combining computational modeling with cognitive science to understand how both systems see and understand the world.

As a PhD candidate at the Max Planck Research Group Vision and Computational Cognition, supervised by Martin Hebart, I work on interpretable alignment of representations—bridging cognitive science and machine learning in both directions. I examine similarities and differences between human and artificial systems, primarily in vision but drawing from language and other domains.

2018–2021
MSc in Artificial Intelligence Radboud University Excellent with distinction, top 1%
2013–2017
BSc in Cognitive Science Osnabrück University Excellent
2025
Nature Machine Intelligence
Florian P. Mahner, Lukas Muttenthaler, Umut Güçlü, Martin N. Hebart
2025
Communications Psychology
Johannes Roth, Yunyan Duan, Florian P. Mahner, Philipp Kaniuth, Thomas SA Wallis, Martin N. Hebart
2025
eLife
Philipp Kaniuth, Florian P. Mahner, Jonas Perkuhn, Martin N. Hebart
Python · 174 stars · 23 forks
A Python toolbox for extracting and analyzing deep neural network representations
Python · 22 stars · 3 forks
A simple TOML configuration parser for Python projects
Python · 12 stars · 3 forks
Code for analyzing object dimensions in humans and DNNs
Python · 4 stars
PyTorch implementation of multiple testing corrections

View all on GitHub →