Erik Matias Lintunen

I'm an incoming PhD student at the University of Alberta, working with Marlos C. Machado on representation-driven discovery of temporal abstractions for reinforcement learning. I'm part of the REAL and RLAI labs, Amii, and the IMOL community. Previously, I spent several years at Aalto University: with Christian Guckelsberger, I studied intrinsically motivated skill learning; with Antti Oulasvirta, how individual differences in cognition predict everyday digital skills.

How do humans and machines make sense of the big and noisy world? I want to better understand this by studying the computational mechanisms that support perception, cognition, and action. To that end, I am drawn to representation learning, the mechanistic interpretability of neural networks and of their learning dynamics, and the interplay between deep learning and reinforcement learning.

Getting here has been wonderfully non-linear: I have formal training in sound art, information experience design, mathematics, statistics, and computational methods. Along the way, I've worked as a sound engineer in the performing arts, a technical producer on large-scale radio broadcasting projects, a software engineer, and a lecturer in digital arts and computing.