Postdoctoral Position in Generative AI and Simulation-supervised Machine Learning for Biology University of Copenhagen in Denmark

Postdoctoral Position in Generative AI and Simulation-supervised Machine Learning for Biology at the University of Copenhagen, Department of Computer Science (DIKU), within the IMAGE section.

This postdoctoral project is part of the Differentiable Realism from AI and Modeling (DREAM) project, funded by the Novo Nordisk Foundation. The research aims to develop self-supervised and unsupervised machine-learning methods for biological data where labels are scarce, expensive, or unavailable. The project combines unlabelled experimental data with differentiable simulations to create realistic synthetic datasets, transferable representations, and robust downstream models without relying on large human-labelled datasets.

Possible directions include generative distribution matching, domain adaptation, diffusion- or score-based models for noise and style transfer, contrastive and masked self-supervised learning, uncertainty-aware learning, and few-shot adaptation. The work is rooted in scientific machine learning, with biological imaging and biophysical simulation as the main testbeds, and datasets ranging from chromatin organization to microorganism tracking.

The position is hosted by the IMAGE section, which conducts research in machine learning, computer vision, and simulation, within the Faculty of SCIENCE. The postdoc is expected to work independently, develop research software, and collaborate with researchers on application-specific projects and experimental validation data.

Applicants should have a PhD or equivalent research experience in computer science, machine learning, scientific computing, physics, applied mathematics, engineering, computational biology, or a closely related area. Experience with generative AI, self-supervised learning, differentiable programming, JAX or PyTorch, scientific machine learning, inverse problems, differentiable simulation, computer vision, or biological image analysis is especially relevant.

The appointment is fixed-term for 1 or 2 years. The expected start date is 1 January 2027, or as soon as possible thereafter. Salary starts at 39,900 DKK per month (approx. 5,300 EUR, August 2026 level), with pension contributions and possible supplements depending on qualifications.

The application deadline is 15 September 2026, 23:59 CET. Applications must be submitted in English through the online application system and should include a motivation letter, CV, diplomas and transcripts, a selected publication/project/code repository, thesis material if available, reference contacts, publication list, and reference letters if available.