I obtained my PhD in machine learning in November 2025 from Linköping University where I was supervised by Professor Fredrik Lindsten. My thesis concerned deep learning methods with the discovery of new materials being the main motivation. This resulted in works related to graph neural networks and diffusion models published at NeurIPS, ICML, and AISTATS. Currently, I am focused on generative models.

I am on the job market: I will for now stay at Linköping University, but I am still open for new opportunities.

Latest News

• 2025-11-07 I have defended my thesis! 🎓🎓
• 2025-10-20 DDSMC is now implemented in the InverseBench codebase
• 2025-10-16 Visited LST4 at EPFL and talked about sequential Monte Carlo and combining it with deep generative models.
• 2025-10-06 My PhD thesis is now available online! I’ll defend on November 7.
• 2025-07-03 Code for DDSMC is now out: https://github.com/filipekstrm/ddsmc
• 2025-05-01 Both WyckoffDiff and DDSMC were accepted to ICML! See you in Vancouver! 🚀🚀🚀🚀
• 2025-04-30 Code for our generative model for materials, WyckoffDiff, is now out: https://github.com/httk/wyckoffdiff