Profile

Lukas is a full-time Senior Research Scientist at Aignostics and a part-time Senior Researcher in the Explainable Machine Learning Group at TUM, led by Prof. Zeynep Akata. Prior to that, Lukas did his PhD in the Machine Learning Group at Technische Universität Berlin, led by Prof. Klaus-Robert Müller. Throughout most of his PhD Lukas was a Student Researcher at Google DeepMind (legacy Brain). At Google DeepMind Lukas was mainly advised by Simon Kornblith, Andrew Lampinen, Thomas Unterthiner, and Klaus Greff. Before his PhD, Lukas did a Master’s in IT and Cognition with a minor in Computer Science at the University of Copenhagen where he mostly worked on Natural Language Processing.

Lukas’ research mainly revolves around (self-supervised) representation learning in computer vision. He works on representational alignment at the intersection of Machine Learning, Cognitive Science, and Medicine, and is driven by his interest in taking inspiration from human cognition to build robust machine learning algorithms.

Publications

"Aligning Machine and Human Visual Representations across Abstraction Levels"
Under review at Nature
Paper

"Getting aligned on representational alignment"
Under review at TMLR
Paper

"When Does Perceptual Alignment Benefit Vision Representations?"
NeurIPS 2024
Paper

"Improving neural network representations using human similarity judgments"
NeurIPS 2023
Paper

"Human alignment of neural network representations"
ICLR 2023
Paper