Ioana Popescu
MSc Student in Data Science · ETH Zurich
Diffusion & Flow Matching · Foundation Models & Transformers · ML for Multi-omics
About
I am Ioana Popescu, a second-year Master's student in Data Science at ETH Zurich and an ESOP Excellence Scholar. My coursework spans statistics, optimization, classical and advanced machine learning, and ML for genomics and medical imaging. Before that, I completed a dual degree in Mathematics and Computer Science at École Polytechnique in Paris, where I finished valedictorian.
My interests include:
- Diffusion & flow matching: sparked by my research with Prof. Valentina Boeva on generative models for spatial transcriptomics, and by the Advanced Machine Learning course at ETH.
- Transformers & foundation models: theoretically through our ICML 2026 paper on the expressivity of transformers for in-context learning, and practically through the ML for Genomics course, where I saw downstream uses of single-cell foundation models.
- Bayesian & active learning: rooted in the Probabilistic AI course at ETH, which I first took and later TAed.
- ML for multi-omics: developed through the Boeva lab project on 3D spatial transcriptomics and ongoing lab meetings on methods for multi-omics data.
I previously interned as a Software Engineer at Microsoft, and this summer as a Quantitative Research Intern at Optiver. I am looking for PhD positions in theoretical and statistical machine learning as well as ML for genomics, and for opportunities in the biotech industry.
Outside of academics, I enjoy reading french literature, playing video games, cooking, and pilates.
News
-
Sep 2026 Completed a Quantitative Research Internship at Optiver.
-
Jul 2026 One paper "Efficient and Minimax Optimal In-context Nonparametric Regression with Transformers" accepted at ICML 2026.
-
Jun 2026 Completed a research project on reconstructing 3D spatial transcriptomics volumes with Prof. Valentina Boeva's group at ETH Zurich.
-
Sep 2025 Started my MSc in Data Science at ETH Zurich with an ESOP Excellence Scholarship from the ETH Foundation.
-
Aug 2025 Special mention from the jury of the Junior Maryam Mirzakhani Prize 2025 for my bachelor thesis on the energy of superlinear preferential attachment trees.
-
Jul 2025 Research internship at the University of Cambridge on the expressivity of transformers for in-context learning.
-
Jun 2025 Graduated valedictorian from École Polytechnique (Mathematics and Computer Science).
-
Jul 2024 Software Engineer Intern at Microsoft.
-
Jun 2024 Received the Women in Computer Science Scholarship from Bending Spoons.
Research Highlights
-
Efficient and Minimax Optimal In-context Nonparametric Regression with Transformers
- At the University of Cambridge, under the supervision of Prof. Richard Samworth, Tianyi Ma, and Dr. William Underwood, we studied the expressivity of transformers for in-context learning and showed that a pretrained transformer with few parameters and pretraining sequences is minimax-optimal for ICL regression.
-
Reconstruction of 3D Spatial Transcriptomics Volumes
- Developed a generative model based on unbalanced optimal transport and thin-plate splines to reconstruct missing slices in 3D spatial transcriptomics volumes, and evaluated it against interpolation-based and flow-matching baselines.
Selected Awards
- ESOP Excellence Scholarship, ETH Foundation · 2025
- "Mathématique au féminin" Scholarship, FMJH · 2024
- Excellence Scholarship, Fondation de l'X · 2022
- Romanian Mathematical Olympiad (RMO): Gold Medal (ranked 1st, max score) · 2022
- European Girls' Mathematical Olympiad (EGMO): Gold Medal (ranked 5th in Europe) · 2021