"PhD Position: Theory of Learning in Artificial and Biologically Inspired Neural Networks"

Are you fascinated by the incredible capabilities of neural networks? Are you interested in applying theoretical methods to understand computational efficiency in neural systems? If so, come and join us as a PhD candidate. You will work in a highly interdisciplinary group, at the intersection of physics, machine learning and theoretical neuroscience.

Our group is focused on investigating dynamics and learning in artificial and biological neural networks, with the aim of:

• Unveiling the link between network structure and neural representations.
• Understanding the impact of structural and energetic constraints on circuit function.
• Deriving design principles for neural networks performing complex tasks.

We employ computational and analytical methods from applied mathematics and physics – particularly statistical mechanics of disordered systems – to study fundamental computational limits of neural networks and their relation to structural and biological constraints in neural systems.

As a PhD candidate, you will investigate the learning capabilities of feed-forward and recurrent models of neural circuits with various degrees of biological plausibility, with a focus on:

• Transferability of representations in multi-task settings.
• The role of biological constraints and network heterogeneity in learning.
• Learning efficiency from an information-theoretic and energetic perspective.

You will have the opportunity to collaborate with other PhD candidates in the lab and international collaborators on both theoretical and computational projects. You will help supervise Bachelor’s and Master’s students, as well as collaborate in the teaching and tutorial sessions in courses on introductory and advanced Machine Learning within the Neurophysics Master’s degree programme at Radboud University. Your teaching load may be up to 10% of your working time.

Requirements:

• You hold an MSc in physics, engineering physics or mathematics.
• You have a good command of analytical techniques used in the modelling of complex systems, and of statistical mechanics methods (particularly as used in the physics of spin glasses). Previous exposure to control theory is a plus.
• You are highly motivated and curious to explore novel research directions in machine learning and theoretical neuroscience.
• You are ready to engage in team work and collaborations with other PhD candidates, both in the group and internationally.
• You have a genuine multidisciplinary interest in the intersection of machine learning and neuroscience.
• You have a good command of spoken and written English.
• You have experience in programming (e.g. Python, C, Julia). The ability to run large-scale simulations on an HPC cluster is a plus.

Conditions of employment:

We will give you a temporary employment contract (1.0 FTE) of 1.5 years, after which your performance will be evaluated. If the evaluation is positive, your contract will be extended by 2.5 years (4-year contract).

You will receive a starting salary of €3,059 gross per month based on a 38-hour working week, which will increase to €3,881 in the fourth year (salary scale P).

You will receive an 8% holiday allowance and an 8.3% end-of-year bonus.