Florian Barkmann
I am a PhD student in the Boeva Lab at ETH Zürich. Previously, I worked as a research intern at the German Cancer Research Center (DKFZ) on optimization problems in radiation therapy and the German Climate Computing Center (DKRZ) developing tools for climate simulations. I hold a Master’s degree in Data Science from ETH Zürich and Bachelor degrees in Mathematics and International Economics from the University of Tübingen. My research interests lie at the intersection of single cell foundation models, representation learning, and cancer biology.
news
| Mar 03, 2026 | Our paper “Sparse Autoencoders Reveal Interpretable Features in Single-Cell Foundation Models” with Flavia Pedrocchi was accepted at the Gen^2 workshop @ ICLR. |
|---|---|
| Feb 16, 2026 | Our review paper “Representation learning of single-cell RNA-seq data” with Constantin Ahlmann-Eltze and Jan Lause was published in RNA. |
| Nov 22, 2025 | Our paper “CanSig Benchmarks Methods for Reproducible Cancer Cell State Discovery from Single-Cell Transcriptomic Data” was accepted in Cancer Research. This work was co-first authored with Josephine Yates. |
| Oct 01, 2025 | We released our preprint “Sparse Autoencoders Reveal Interpretable Features in Single-Cell Foundation Models” with Flavia Pedrocchi. We trained sparse autoencoders on hidden representations of three single-cell foundation models to study their internal features. |
| Sep 09, 2025 | I joined Cradle in Zurich as a machine learning research intern. |