curriculum vitae
Full Name | Lukas Florian Münzel |
Languages | English, German (native), Allemanic (native), French (B2) |
Education
-
2022 - 2025
Bachelor of Science in Mathematics
ETH Zürich
-
Current grade average: 5.8 / 6.0. Spent a wonderful spring semester 2025 on exchange at UC Berkeley
- Functional Analysis I
- Information Theory I
- Markov Processes
- Introductory Mandarin
- Seminar on Statistical-to-Computational Gaps
- Probability Theory
- Fundamentals of Mathematical Statistics
- Dynamical Systems and Ergodic Theory
- Fourier Theory and Hilbert Spaces
- Mathematics of Signals, Systems, and Learning
- Probability and Statistics
- Topology
- Deep Learning (graduate level)
- Measure Theory
- Algebra I
- Complex Analysis
- Numerical Analysis I
- Linear Algebra II
- Real Analysis II
- Models of Computation
- Group theory, Zermelo–Fraenkel set theory and logic
- Chemistry (audited)
- Introduction to Biology II (audited)
- Linear Algebra I
- Real Analysis I
- Neuromorphic Engineering I (graduate level, dropped after one month)
- Physics I
- Introduction to programming
Coursework Fall 2025
Coursework Spring 2024
Coursework Fall 2023
Coursework Spring 2023
Coursework Fall 2022
-
2022
Maturitätszeugnis / Swiss high school diploma
Gymnasium Bäumlihof, Basel
- Founded and lead a club in which motivated students could learn competitive programing and computational thinking
- Skipped two grades
- Graduated with highest praise and a grade average of 5.55 / 6.0, receiving the "Basler Maturapreis"
Research experience
-
May 2025
University of California, Berkeley
Independent work from exchange semester
- Proved optimality of the KL divergence estimator introduced by John Schulman that is used in PPO and GRPO
- Blogpost: A small observation on Schulmann's KL divergence estimator
-
Sep 2023 - Jan 2024
ETH Zürich
Project as part of graduate-level course on Deep Learning
- Worked on enhancing representations of heterophilic graphs, a type of graph with which Graph Neural Networks commonly struggle
- I implemented and ran various experiments with PyTorch Lightening to investigate the performance of our proposed modifications. With these modifications, we surpassed the state-of-the-art accuracy of 91.64% by over three percent on the heterophilic dataset Minesweeper
-
Jul - Aug 2021
Massachusetts Institute of Technology
Only central European student selected for the Research Science Institute 2021
- Designed, programmed and conducted experiments to demonstrate the feasibility of a supplemental approach to training artificial neural networks under supervision of Prof. Lizhong Zheng
- Click here for a presentation recording and here for the report
-
Jan - Mar 2021
ETH Zürich
Eight week full-time internship under supervision of Prof. Patrick Jenny
- Optimized and implemented code for numerical fluid simulations in order to simulate airflow in lecture halls
- Designed and implemented a novel tool to edit the simulation meshes
- This work is in the process of being published with me as third author
Honors and Awards
-
2024
- Awarded membership to the Swiss Study Foundation
- Invited to hold a talk on "Progress, machine learning, and why to be optimistic about the future" at the Zurich Undergraduate Colloquium in Physics and Mathematics
-
2022
- Gold medal and second place at the final round of the Swiss Olympiad in Informatics, representing Switzerland at the International Olympiad in Informatics in Indonesia
- Basler Maturapreis by Novartis. Selected for this price for ”outstanding dedication and accomplishments, socially and academically” out of a class of 111 students
- 15th place nationally in the second round, participating in the week-long final round involving various theoretical and laboratory exams
-
2020
- Second place at the final round of the Swiss Olympiad in Informatics and therefore qualifying for the International Olympiad in Informatics held virtually by Singapore