Get Started
Home
Topics
Search
Library
Research questionHow can node-level over-smoothing diagnostics guide virtual-node selection to alleviate over-squashing in GNNs?Deep GNNs can make node representations indistinguishable, while long-range information may be compressed through limited message-passing paths. Global measures can miss heterogeneous node behavior, making it difficult to identify which nodes should support information flow.
AI
Machine Learning
Research Paper
Technology
Latest papersRecent research connected to this question, newest first.LEED: Local Embedding Evolution Distance for over-smoothing estimation and virtual node selection in GNNThe source studies GNNs using a local embedding-evolution metric for node-level over-smoothing analysis and virtual-node selection. It reports comparisons with Dirichlet energy and experiments on multiple datasets, but does not establish broader deployment or access requirements.research paper · Sep 4, 2026
Related questions
How can graph neural networks capture long-range interactions without oversmoothing or oversquashing on diverse large graphs?How can reservoir-based graph neural networks aggregate multi-hop information without over-smoothing?How should graph representations be chosen for GNN-based reinforcement learning in power-grid topology control?How can graph learning leverage visual graph depictions for structural reasoning?