About me

For the first half of 2026, I am a visiting researcher at the INDY lab at EPFL, working with Matthias Grossglauser and Patrick Thiran.

I am an FCAI postdoctoral researcher at the University of Helsinki, working within the research groups of Arto Klami and Luigi Acerbi. Previously, I completed my PhD at Aalto University under the supervision of Samuel Kaski. I also hold two Master’s degrees, one in Mathematics and one in Economics, from the University of Helsinki.

Research

My research focuses on probabilistic deep learning, particularly on learning from human data. At its core, my work combines random utility theory with modern probabilistic models—such as flows, diffusion models, and autoregressive transformers—to create accurate representations of the utilities and beliefs held by various oracles, including humans and LLMs. As an example, see work on normalizing flows and diffusion models. Key areas of interest include probabilistic deep learning, choice modeling, reward modeling, elicitation (knowledge/prior elicitation, preference learning), and Bayesian optimization.