Foundation Models for Combinatorial Search

About the role

The position is a research role within Prof. Siegfried Nijssen’s lab at KU Leuven. The project aims to develop foundation models that can solve combinatorial search problems by combining neural networks with search algorithms.

Key responsibilities

• Conduct theoretical and applied research on neural‑network‑guided combinatorial search.
• Design and implement neural components (e.g., transformers, CNNs) that steer search procedures.
• Develop and evaluate search algorithms such as branch‑and‑bound or beam search on increasingly large problem classes.
• Collaborate with the DTAI research group and contribute to the Flanders AI programme.
• Publish findings in top‑tier conferences and journals.

Required profile

• Excellent Master’s degree in Computer Science, Artificial Intelligence or a related discipline.
• Strong background in both the theory and implementation of neural networks and/or combinatorial search algorithms.
• Demonstrated programming expertise.

Required skills

• Transformers
• Convolutional Neural Networks (CNNs)
• Branch‑and‑bound search
• Beam search
• Programming (e.g., Python, C++)

What we offer

• Opportunity to work in a leading AI research group.
• Access to funding through the Flanders AI programme.
• Collaborative, interdisciplinary research environment.