Project DTRIP4H partner University of Helsinki (Finland) participated in the NeurIPS 2025 conference held in the United States, contributing to the workshop “New Perspectives in Graph Machine Learning.” The workshop took place in San Diego on December 7, 2025.

During the workshop’s poster session, university researchers presented work related to the DTRIP4H project through the publication “Exploring Heterophily in Graph-level Tasks.” The study establishes a new theoretical understanding of heterophily in graph-level learning and offers guidance for designing effective GNN architectures.

Participation in one of the world’s leading conferences in artificial intelligence and machine learning provided an opportunity to share the project’s research results with the global scientific community, engage with leading experts in the field, and discuss future directions for graph-based machine learning methods. More on the conference: https://neurips.cc/

The presentation highlighted the ongoing contributions of the DTRIP4H project and its partners to advancing research at the intersection of AI, data science, and complex network analysis.

This poster is part of the publication, available open-access on Cornell University preprint repository arXiv: https://arxiv.org/abs/2509.18893