M34.1 CONNECT THE MECHANISM
Learn by exchanging messages across a graph
A molecule has no first atom and no fixed size. Watch atoms pass messages along their bonds, by hand, and see why too many rounds blur every atom into the same thing.
LESSON OVERVIEW14 min lesson
Lesson overview
A molecule has no first atom and no fixed size. Watch atoms pass messages along their bonds, by hand, and see why too many rounds blur every atom into the same thing.
What you’ll explore
- Graph neural networks update node or edge representations using neighborhood messages and aggregation; graph construction, permutation behavior, and depth determine what relationships can be represented.
GO TO THE SOURCE
Original explanations, connected to the research.
Neural Message Passing for Quantum Chemistry (Gilmer et al., 2017)Semi-Supervised Classification with Graph Convolutional Networks (Kipf & Welling, ICLR 2017)Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning (Li, Han & Wu, 2018)How Powerful are Graph Neural Networks? (Xu et al., ICLR 2019)On the Bottleneck of Graph Neural Networks and its Practical Implications (Alon & Yahav, ICLR 2021)Suggest a correction
A precise note can make an explanation better.
Choose the scene and describe what needs attention. Download a feedback file to share through a channel you already use. This page does not send feedback or connect you with a reviewer.