Learning on Graphs Conference 2026

Learning on Graphs Conference (LoG) 2026 is scheduled for 20–22 November 2026 in Boston, the second in-person edition of this fast-rising research conference. LoG covers machine learning on graphs and geometry — graph neural networks, geometric deep learning, and relational/structured learning — and is distinctive for an explicit emphasis on review quality, even offering monetary reviewer rewards. It also runs a worldwide network of local meetups (Pisa, Tromsø, Gandhinagar, New Delhi, Paris, São Carlos) to broaden access.

Graph and geometric methods underpin everything from molecular and materials discovery to recommendation and knowledge graphs, making LoG an important specialist anchor as these techniques fuse with foundation models.