PhD Candidate: Generative Diffusion Modelling and Associative Memory at the Donders Centre for Cognition - Hiring in process/Finished, not possible to apply

1. Introduction

Are you interested in generative diffusion models and how they connect to memory? Come and join the Generative Memory Lab (DCC) as a PhD candidate under the supervision of Dr Luca Ambrogioni. We carry out research at the frontier of generative modelling using concepts from statistical physics and neuroscience. For example, we have shown that the phenomenon of symmetry breaking is central to generative diffusion. If you want to be a part of this fast-changing field while doing rigorous science, this is the position for you.

2. Job description

As a PhD candidate, your main responsibility will be to work on research projects in generative machine learning. These projects require both training of deep neural networks on large datasets and performing theoretical analyses and calculations to elucidate their properties. In particular, you will become proficient in training and analysing state-of-the-art diffusion models and variants such as the newly introduced consistency models. These models rest on a joint foundation of deep learning architectural research and mathematical theory. They therefore present an exciting research area for theoretically minded individuals who are also interested in the practice and applicability of state-of-the-art generative modelling.

Your research will focus on the interconnections between generalisation and memorisation in generative models. The aim is to understand the ‘magic’ of generative diffusion models and cast light on the computational basis of our own human memory as well. Depending on the specific project, you will collaborate both with statistical physicists and with experimental and theoretical neuroscientists, such as Dr Freyja Ólafsdóttir (DCN).

As part of your duties (10%) you will assist in teaching courses in machine learning and mathematics for Bachelor’s and Master’s students. You will also co-supervise ambitious Master’s students with their research projects.

3. Profile

• You have a Master’s degree in Computer Science, Mathematics, Physics, Statistics, Artificial Intelligence, Machine learning or another relevant field.
• You have a strong background in machine learning and deep learning.
• You are proficient in Python scientific programming and at least in one deep learning framework such as Pytorch, Jax or TensorFlow.
• You are ambitious and willing to work in a fast-paced and competitive research environment.
• Demonstrable experience in training diffusion models is preferred.
• A background in statistical physics is also preferred.
• Familiarity with basic concepts of the theoretical neuroscience of memory, such as Hopfield networks, is desirable.

4. We are

The Donders Institute for Brain, Cognition and Behaviour is a world-class interfaculty research centre that houses more than 700 researchers devoted to understanding the mechanistic underpinnings of the human mind. Research at the Donders Institute is focused around four themes: 1. Language and communication, 2. Perception, action and control, 3. Plasticity and memory, 4. Neural computation and neurotechnology. Excellent, state-of-the-art research facilities are available for the broad range of neuroscience research that is being conducted at the Donders Institute. The Donders Institute has been assessed by an international evaluation committee as ‘excellent’ and recognised as a ‘very stimulating environment for top researchers, as well as for young talent’. The Donders Institute fosters a collaborative, multidisciplinary, supportive research environment with a diverse international staff. English is the lingua franca at the Institute.

In the Generative Memory Lab, we work on understanding and exploiting these phenomena by combining generative machine learning methods with ideas from neuroscience, probability theory and theoretical physics. We aim at both advancing the state-of-the-art of generative artificial intelligence and using these insights to cast light on the nature of the continuum connecting human memory and imagination. In addition to this more theoretical work, we conduct applied research in probabilistic inference, deep learning and medical prediction and decision making.

5. We offer

• We will give you a temporary employment contract (0.8 FTE 5- year contract - 1.0 FTE 4- year contract) 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) or 3.5 years (5-year contract).
• You will receive a starting salary of €2,770 gross per month based on a 38-hour working week, which will increase to €3,539 from the fourth year onwards (salary scale P).
• You will receive an 8% holiday allowance and an 8,3% end-of-year bonus. 
• You will be able to use our Dual Career and Family Support Service. The Dual Career Programme assists your partner via support, tools, and resources to improve their chances of independently finding employment in the Netherlands. Our Family Support Service helps you and your partner feel welcome and at home by providing customised assistance in navigating local facilities, schools, and amenities. Also take a look at our support for international staff page to discover all our services for international employees.
• You will receive extra days off. With full-time employment, you can choose between 30 or 41 days of annual leave instead of the statutory 20. 

6. Practical information and applying

You can apply no later than 27 May only via the button below. Address your letter of application to Luca Ambrogioni. In the application form, you will find which documents you need to include with your application.

The first interviews will take place on 17 June. Any second interview will take place on 24 June. You will preferably start your employment on 1 September.

We can imagine you're curious about our application procedure. It describes what you can expect during the application procedure and how we handle your personal data and internal and external candidates.