Studying permutation-invariant graph representation learning and graph generation
Title: Studying permutation-invariant graph representation learning and graph generation
DNr: Berzelius-2025-111
Project Type: LiU Berzelius
Principal Investigator: Roman Bresson <bresson@kth.se>
Affiliation: Kungliga Tekniska högskolan
Duration: 2025-03-21 – 2025-10-01
Classification: 10210
Keywords:

Abstract

I intend to implement and run graph learning models such as: -Graph autoencoders -Graph transformers as well as generative methods such as: -Flow matching -Diffusion. The aim is to study and improve graph generation paradigms, especially through the studies of permutation-invariant generation.