Postdoc position in Probabilistic Machine Learning and Amortized Inference

The Department of Computer Science, Faculty of Science, University of Helsinki invites applications for a Postdoctoral Researcher in Probabilistic Machine Learning and Amortized Inference.

The University of Helsinki is an international scientific community of 40,000 students and researchers. It is one of the leading multidisciplinary research universities in Europe and ranks among the top 100 international universities in the world.

The Department of Computer Science, which is part of the Faculty of Science, is a leading Computer Science research and teaching unit in Finland. The research themes of the Department cover machine learning and algorithms, computer networks and distributed systems, software systems and bioinformatics.

The Department of Computer Science, Faculty of Science, University of Helsinki invites applications for a

Postdoctoral Researcher in Probabilistic Machine Learning and Amortized Inference

Position description

The Machine and Human Intelligence research group led by Principal Investigator Luigi Acerbi is looking for a postdoctoral researcher eager to work on new machine learning methods for robust, efficient, surrogate-based, simulator-based, and amortized probabilistic inference with applications in fields such as computational and cognitive neuroscience. The candidate will join our highly collaborative research group at the Department of Computer Science of the University of Helsinki with strong links to the Finnish Center for Artificial Intelligence FCAI.

Recent advances in machine learning have shown how powerful emulators and surrogate models can be trained to drastically reduce the costs of simulation, optimization and Bayesian inference, with many trailblazing applications in the sciences. In this project, the candidate will join an active area of research within the group to develop new methods for simulation, optimization and inference that combine state-of-the-art deep learning and surrogate-based approaches – including for example deep sets and transformers; normalizing flows; Gaussian and neural processes – with the goal of achieving maximal sample-efficiency (in terms of number of required model evaluations or simulations) and wall-clock speed at runtime (via amortization). The candidate will apply these methods to challenging problems involving statistical and simulator-based models that push the current state-of-the-art, be it for number of parameters (high-dimensional amortized inference), number of available model evaluations (extreme sample-efficiency), amount of data, or by tackling open questions such as generalization and robustness to model misspecification. The project will build on the state-of-the-art frameworks developed in our group, from Variational Bayesian Monte Carlo (VBMC; Acerbi, NeurIPS; 2018, 2020) to the recently proposed Amortized Conditioning Engine (Chang et al., 2024). Example applications include state-of-the-art computational cognitive models of sequential decision making from our collaborators (van Opheusden et al., Nature; 2023).

The project will include research visits abroad to work with collaborators, such as Wei Ji Ma's group at New York University (New York, NY, USA), Alexandre Pouget's group at the University of Geneva (Switzerland) and The International Brain Laboratory and Valentin Wyart at the Ecole Normal Supérieure in Paris (France). The project will also include multiple opportunities for collaboration with the thriving machine learning community of the Finnish Center for Artificial Intelligence (FCAI).

Expected milestones include writing and submitting articles to leading machine learning and statistics venues, such as NeurIPS, ICML, AISTATS, or established computational journals such as PLoS Computational Biology.

The position is full-time, funded for two years possibly renewable (pending funding), with a target starting date in early 2025.

Requirements

The work involves research-related activities (~95% of workload), including: conducting theoretical and applied research, collaborating and interacting with team members, designing and programming machine learning software, computational and data analyses, writing research articles, participating in and presenting research at academic conferences, and taking part in international research visits. Teaching-related activities comprise approximately 5% of the workload.

The ideal candidate has a strong background in computational statistics and/or machine learning, particularly in approximate Bayesian inference and/or probabilistic machine learning methods (e.g., MCMC, normalizing flows, variational inference, Bayesian deep learning, etc.). We also require experience in implementing efficient code in Python (e.g., PyTorch, Jax). The following constitute an advantage, but are not required:

  • Experience with Gaussian processes and/or Bayesian optimization. 
  • Experience with simulator-based inference and/or amortized inference methods and toolboxes.
  • Previous research experience and publications (required for a postdoctoral position).

Excellent written and oral communication skills in English are needed.

Applicants should have – or be about to obtain – a PhD in machine learning, computer science, data science, applied mathematics and statistics, physics, or a related field.

Salary and benefits

The position is full-time and funded for 2 years (postdoc). The starting salary is 2400–2800 euros/month (for a doctoral student) and 3600–4200 euros/month (for a postdoc), depending on previous qualifications and experience.

The University of Helsinki offers comprehensive services to its employees, including occupational health care and health insurance, sports facilities, and opportunities for professional development. The International Staff Services office assists employees from abroad with their transition to work and live in Finland.

How to apply

Please submit your application with its attachments to the group leader, Luigi Acerbi (please find email address below), specifying "Application for postdoc in Probabilistic Machine Learning and Amortized Inference" as the email subject.

The application should include the following attachments as pdf files (in English):

  1. CV with possible list of publications and previous projects;
  2. A copy of your transcripts (list of courses completed during BSc/MSc and grades);
  3. Cover letter (1-2 pages) with motivation, research interests and match to the project;
  4. Contact details of two referees who could provide a letter of recommendation (can be included at the end of the CV);
  5. An extended sample of your scientific writing (e.g., your Master's or PhD thesis, or a project report; could be a link if available online).

Applications will be considered until the position is filled.

Additional information

For more information on the group and the position, please:

And these are some good reasons to move to Finland.

Our group is committed to principles of diversity, equality and inclusion. We encourage applications from women, racial and ethnic minorities, and other under-represented individuals.

 

Appendix: Background references and material

Surrogate-based probabilistic machine learning methods developed in our group:

  • Acerbi (2018). Variational Bayesian Monte Carlo. NeurIPS (link).
  • Acerbi (2020). Variational Bayesian Monte Carlo with Noisy Likelihoods. NeurIPS (link).
  • Li, Clarté, Jørgensen & Acerbi (2024). Fast post-process Bayesian inference with Variational Sparse Bayesian Quadrature. arXiv preprint (link).

Amortized Inference work from our group and in collaboration with FCAI:

  • Huang, Haussmann, Remes, John, Clarté, Luck, Kaski & Acerbi (2023). Practical Equivariances via Relational Conditional Neural Processes. NeurIPS (link).
  • Huang, Bharti, Souza, Acerbi & Kaski (2023). Learning Robust Statistics for Simulation-based Inference under Model Misspecification. NeurIPS (link).
  • Chang, Loka, Huang, Remes, Kaski & Acerbi (2025). Amortized Probabilistic Conditioning for Optimization, Simulation and Inference. AISTATS (link).