Email | LinkedIn | Google Scholar
I am a PhD fellow at the Machine Learning Section of the University of Copenhagen
with a focus on Unlearning, Robustness and Privacy.
I am supervised by Prof. Amartya Sanyal
and Prof. Amir Yehudayoff and part of the Foundations of
Responsible Machine Learning Group (Cope-FoRML).
|
arxiv 2026
TPDP 2026 |
Less Noise, Same Certificate: Retain Sensitivity for Unlearning | Slides Carolin Heinzler, Kasra Malihi, Amartya Sanyal Certified unlearning differs from differential privacy in that the retained data is fixed, allowing removal to be certified with less noise than would be required under worst-case DP sensitivity. |
|
SODA 2026
FORC 2026 |
Learning in an Echo Chamber: Online Learning with Replay Adversary Daniil Dmitriev, Harald Eskelund Franck, Carolin Heinzler, Amartya Sanyal$^*$ | $^*\alpha\beta$-cal order | Slides When models train on their own past guesses, mistakes can echo and mislead learning. We introduce a learning-theoretic setting that models this phenomenon: Online Learning in the Replay Setting. We introduce a combinatorial measure, the Extended Threshold dimension, which characterises learnability in this setting. |
|
Master's thesis
arxiv 2024 |
Adversarial Resilience against Clean-Label Attacks in Realizable and Noisy Settings Carolin Heinzler We investigate the challenge of establishing stochastic-like guarantees when learning from a stream of i.i.d. data with clean-label adversarial samples. Introducing the notion of a clean-label adversary in the agnostic context, we are the first to give a theoretical analysis of a disagreement-based learner for thresholds. |
| Reviewer | EurIPS 2025 Workshop PPML, NeurIPS 2025 Workshop Reliable ML, AISTATS 2025 |
| P1 Program | Member of P1 Program: Data Privacy in Machine Learning of the Pioneer Center for AI, Denmark |
| Local Organizer |
Affinity Event of the Learning Theory Alliance at EurIPS 2025 in Copenhagen, Denmark IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) 2025 in Copenhagen, Denmark |
| Representation | PhD student representative in LSU committee for Computer Science Department, University of Copenhagen (2024-ongoing) |
| Teaching Assistant |
Machine Learning A (2025, University of Copenhagen), Mathematics of Signals, Networks and Learning (2024, ETH Zurich), Quantitative Risk Management (2024, ETH Zurich), Probability Theory and Statistics (2023, ETH Zurich), Introduction to Mathematics (2021-2022, WU Vienna), Introduction to Phyton (2019-2021, University of Vienna) |