PhD in Probabilistic and Differential Algorithms: Utrecht University is offering an exciting PhD opportunity as part of the ERC project FoRECAST. The position is focused on the development of probabilistic and differential inference algorithms and their applications in machine learning. This is a remarkable opportunity for early-career researchers to dive into foundational research in an interdisciplinary and diverse team.
PhD Position in Probabilistic and Differential Algorithms
Designation
PhD Candidate in Probabilistic and Differential Algorithms
Research Areas
- Probabilistic and Differentiable Algorithms for Machine Learning
- Programming Language Implementation for High-Performance Computing
- Programming Language Semantics and Foundations
Details of the Position
| Field | Details |
|---|---|
| Institution | Utrecht University |
| Department | Department of Information and Computing Sciences |
| Faculty | Faculty of Science |
| Position Type | Full-time (36 to 40 hours per week) |
| Duration | Initial 18 months with potential extension to 4 years upon successful assessment |
| Salary | €3,059 – €3,881 per month (scale P under the Dutch Universities’ Collective Labour Agreement) |
| Benefits | 8% holiday pay, 8.3% year-end bonus, pension scheme, flexible employment terms, partially paid parental leave, and benefits for professional development, sports, and culture. |
| Application Deadline | 9 September 2025 |
| Location | Utrecht University, Utrecht, Netherlands |
Eligibility/Qualification
Candidates should fulfill the following qualifications:
- A Master’s degree in Computer Science, Mathematics, Statistics, Physics, or a related discipline.
- Strong interest or prior experience in:
- Inference algorithm development and evaluation for statistical computing and machine learning.
- Statistical computing and machine learning applications.
- Probabilistic and differentiable programming.
- Excellent English communication skills (written and verbal).
- Capability to work both independently and collaboratively in a diverse team environment.
- A curiosity-driven approach to research and willingness to define their unique research trajectory within the ERC project.
Description
As a PhD candidate, you will be part of the ERC FoRECAST Project. Your role will focus on area 1: Probabilistic and Differentiable Algorithms for Machine Learning. Some of the tasks include:
- Developing new probabilistic and differential programming techniques, such as gradient estimation for probabilistic programs, implicit function differentiation, and compositional Bayesian inference methods.
- Mathematically proving the correctness and efficiency of these techniques.
- Building high-performance implementations using functional array programming techniques.
- Applying these techniques to tackle real-world problems like experimental design and reinforcement learning.
Research will be conducted under the supervision of Matthijs Vákár and secondary supervisors with expertise in machine learning. Highlights of this opportunity include engagement with an inclusive and inspiring Software Technology group, which hosts experts involved in popular tools like Accelerate and Stan, collaborations with international partners (e.g., University of Oxford and Google JAX teams), and opportunities for teaching.
How to Apply
To apply, submit the following documents via the “apply now” button provided on the Utrecht University website:
- A two-page cover letter addressing:
- Research interests and their relevance to the position.
- Evidence of self-motivation and collaborative ability.
- A summary of your Master’s thesis or a comparable major project.
- Your earliest possible starting date.
- Complete academic CV, including:
- Publications, GitHub repositories (if any), links to blogs, and academic achievements.
- Copies of your degree certificates and transcripts with grades.
- A copy of your MSc thesis (if available).
- Details of two or three references.
Last Date to Apply
9 September 2025
Take advantage of this career-defining opportunity to contribute to the cutting-edge field of probabilistic and differential programming while working with world-class researchers in a collaborative, inclusive setting.
Apply Now!
