Master’s Theses: Neuromorphic and Neuro-inspired AI
We are looking for motivated master’s students to explore cutting-edge topics at the intersection of neuromorphic computing, state-space models (SSMs), transformers, and neuroscience-inspired AI. Specifically, we aim to investigate how neuromorphic principles and the entorhinal-hippocampal formation — the brain’s navigation system — can inspire modern machine learning.
The project can be tailored to align with the student's interests and qualifications.
Motivation
- Energy and data efficiency
- Incorporating inductive biases to improve generalization
Project Area I: Neuromorphic State-Space Models & Transformers
- Neuromorphic principles in deep learning
- State-Space Models (SSMs) and Transformers
- Sparsity, quantization, and algorithmic efficiency
- Potential neuromorphic hardware implementation
Project Area II: Neuroscience-Inspired AI (NeuroAI)
- Learning from the brain’s navigation system
- Grid cells, place cells, and cognitive maps
- Spatial representation and memory in ML
- Novel architectures for learning and planning
Your Profile
- Background in computer science, mathematics, physics, or machine learning
- Strong programming skills (PyTorch, JAX, or similar frameworks)
- Excellent analytical and problem-solving skills; self-motivated and proactive
- Enrollment at RWTH Aachen or willingness to conduct the thesis in Aachen (on-site)
What We Offer
- Cutting-edge research at the intersection of neuroscience and AI
- Opportunity to contribute to novel algorithms and potential scientific publications
- Collaboration with experts in machine learning, neuroscience, and neuromorphic computing
- Flexible research directions (theoretical, algorithmic, or hardware-focused)
- Access to compute resources
Application
Please get in touch with a.renner(at)fz-juelich.de to discuss concrete project ideas.
The preferred project duration is March-August 2025, please mention your preferences.
Contact
Dr. Alpha Renner (a.renner(at)fz-juelich.de)