PhD in Mathematics: Algebraic Geometry and Neural Network Theory at KTH Royal Institute of Technology KTH Royal Institute of Technology in Sweden

PhD position in Mathematics at the intersection of algebraic geometry and neural network theory at KTH Royal Institute of Technology in Stockholm.

The project is supervised by Kathlén Kohn (Associate Professor of Mathematics) and focuses on understanding the double descent phenomenon in deep learning. The research aims to develop proofs using algebraic geometry and to make those proofs constructive, with explicit bounds on model sizes and generalization behavior.

The successful candidate will work in Kathlén Kohn’s research group on Algebraic Geometry in Data Science and AI, and will be part of both the WASP graduate school and KTH’s doctoral program in mathematics. The position is a full-time, 5-year PhD starting in summer or fall 2026.

Funding is provided by WASP – Wallenberg AI, Autonomous Systems and Software Program, with generous travel funds and networking opportunities.

Application deadline: May 20, 2026 (11:59 PM CET). The post says application instructions will be announced on May 7, and interviews are planned for May 25, 2026.

Research interests include algebraic neural network theory, semi-algebraic sets, polynomial networks, loss landscapes, discriminants, singularities, and deep learning theory.