Arsen Sheverdin

Arsen Sheverdin

PhD Student

Helmholtz AI

TU Munich

I embarked on my PhD journey in November 2023 at the Helmholtz AI and the Technical University of Munich, where I am supervised by Dr. Vincent Fortuin. Before this, I completed my research Master’s in Artificial Intelligence at the University of Amsterdam as an ELLIS MSc Honours student.

My primary research goal is to develop intelligent, safe, and robust agents that are capable of (1) not only learning efficiently but also understanding what they do not know, and (2) adapting to changes in data distributions. My main interests encompass probabilistic modeling, deep learning, approximate inference, deep generative models, and information theory. I am also excited about ideas related to imbuing neural networks with inductive biases (e.g., invariance and equivariance) and causality.

Interests
  • Probabilistic modelling
  • Deep generative models
  • Approximate inference and information theory
Education
  • MSc in Artificial Intelligence, 2023

    University of Amsterdam

  • BSc in Physics (Mathematics minor), 2020

    Nazarbayev University

Latest