Summary
The Technische Universität Darmstadt is offering a position for a Research Assistant / Doctoral Candidate (w/m/d) focusing on the mathematical foundations of deep learning within the newly founded Emmy Noether Research Group led by Dr. Mariia Seleznova, under the direction of Prof. Dr. Felix Krahmer.
Designation
Wiss. Mitarbeiter:in / Doktorand:in (w/m/d) (Research Assistant / Doctoral Candidate)
Quick Overview Table
| Host Institution | Technische Universität Darmstadt |
| Department/Area | Signal Processing and Statistical Learning (Fachgebiet Signaltheorie und Statistisches Lernen) / Emmy Noether Research Group |
| Project | RMT4DL (Random Matrix Theory for Deep Learning) |
| Location | Darmstadt, Hesse, Germany |
| Designation | Doctoral Candidate / Research Assistant |
Research Area
The research is part of the project “RMT4DL (Random Matrix Theory for Deep Learning)”. Topics include:
- Methods for deep learning from probability theory and random matrix theory to rigorously analyze stability, generalization, and feature learning in neural networks with a large number of parameters.
- Empirically informed theory of deep learning, identifying and mathematically formalizing empirical phenomena (e.g., scaling laws) and developing theoretical models to explain them.
Location
Darmstadt, Hesse, Germany
Eligibility / Qualification
- Very good university degree (Master or equivalent) in mathematics, computer science, or a closely related field.
- Strong mathematical knowledge, particularly in machine learning theory, probability theory, optimization, and linear algebra. Knowledge in random matrix theory is an advantage.
- Programming experience in Python. Experience with deep learning frameworks (e.g., PyTorch or JAX) is a plus.
- Very good spoken and written English language skills.
- Strong analytical skills and enthusiasm for independent scientific research.
Scholarship / Position Description
As a doctoral candidate, you will conduct research aiming to develop mathematical foundations for modern deep learning. The role serves the purpose of scientific qualification, offering the opportunity to prepare for a doctorate. Tasks include publishing research results in leading journals and/or conferences, presenting at international conferences, and contributing to teaching, mentoring, and administrative duties.
TU Darmstadt offers flexible working models, mobile working, 30 days of vacation per year, 5 days of educational leave, internal training opportunities, comprehensive health and sports offerings, a company pension scheme (VBL), and child care support.
How to Apply
Applications should include a CV, a letter of motivation, transcripts and degree certificates (Bachelor and Master or equivalent), and the names and contact details of one to two academic references. For questions regarding the position, please contact Dr. Mariia Seleznova.
Last Date for Apply
Please check the official application link for current deadline details.
Apply Link
Apply Online at TU Darmstadt Career Portal
