Fabian Falck

About

I am an ML researcher at Microsoft Research Cambridge working on generative models. Before, I did my PhD in Statistical Machine Learning in in the OxCSML group (opens in new tab) at the University of Oxford, supervised by Prof. Chris Holmes (opens in new tab) and Prof. Arnaud Doucet (opens in new tab). I was also a Graduate Teaching and Research Scholar in Computer Science (opens in new tab) at Oriel College, University of Oxford, teaching maths courses, and studied at the The Alan Turing Institute (opens in new tab), the UK’s national institute for AI.

My research focusses on:

• Understanding the inductive biases of generative models, particularly diffusion/flow-matching models (opens in new tab), hierarchical VAEs (opens in new tab), and their neural architecture such as the U-Net (opens in new tab)
• Properties of LLMs, such as their fixed-point behaviour (opens in new tab), counterfactual reasoning (opens in new tab), and their ability to perform Bayesian inference (opens in new tab)
• Large-scale, multi-modal applications of generative models, such as analog-amenable machine learning, inverse problems, and radiology-report generation

Research highlights include a NeurIPS 2022 oral (top 1.76%), an ICML 2025 spotlight (top 2.6%), 2 workshop orals at ICLR and AAAI, a publication in Nature Reviews Genetics and a Best Paper Award at TAROS, the UK’s largest robotics conference before my PhD.

If you would like to collaborate or be co-supervised by me, please reach out!

Contact: fabian.falck@microsoft.com