About
I’m a PhD student in the Statistics and Probability Theory section of the Department of Mathematical Sciences at the University of Copenhagen, where I’m part of the Copenhagen Causality Lab. I’m supervised by Sebastian Weichwald and Anton Rask Lundborg, and I’m also affiliated with the Pioneer Centre for Artificial Intelligence.
Interests: causality, performative prediction.
News
Nov 2026
Our NeurIPS position paper is now on arXiv.
Sep 2026
Our paper Causal discovery needs explicit epistemic standards was accepted to the position paper track at NeurIPS 2026.
Jul 2026
Our paper Two-Stage Causal Discovery with FLOP and DirectLiNGAM was accepted at the Causality for Decision Making workshop at UAI 2026 in Amsterdam. See you there!
Apr 2026
I’ll be at the European Causal Inference Meeting (EuroCIM 2026) in Oxford. Looking forward to seeing you there!
Sep 2025
Started my PhD in the Copenhagen Causality Lab at the University of Copenhagen.
Oct 2024
Finished my M.Sc. in Machine Learning at the University of Tübingen.
Research
2026
NeurIPS 2026 (position)
arXivPDF
Causal discovery needs explicit epistemic standards
Harald Kugler, Leonard Henckel, Sebastian Weichwald
Causal discovery methods are usually compared via identifiability, consistency
and benchmarks, each of which introduces assumptions that are often untestable.
We argue that the epistemic standards for discussing these assumptions should be
made explicit, and that benchmarks should be designed to operationalize them.
2026
UAI 2026 workshop
OpenReviewPDF
Two-Stage Causal Discovery with FLOP and DirectLiNGAM
Harald Kugler, Leonard Henckel, Sebastian Weichwald
FLOP+DL estimates the Markov equivalence class with FLOP and then orients the
edges left undirected with DirectLiNGAM. Accepted at the
Causality for Decision Making
workshop at UAI 2026.
2024
Master's thesis
Code
Towards a reliable and scalable data-copy detection algorithm using random projections
Harald Kugler
Explores detecting data-copying in generative models by comparing generated
samples and training data after random linear projections.
Experience
Sep 2025–now
University of Copenhagen
PhD student
Copenhagen Causality Lab, supervised by Sebastian Weichwald and Anton Rask Lundborg.
Apr–Aug 2025
Technical University of Denmark
Research assistant
DTU Compute, supervised by Line Clemmensen.
Education
2021–2024
M.Sc. Machine Learning
University of Tübingen
2018–2021
B.Sc. Cognitive Science
University of Tübingen