Message-Passing Depth and Over-Smoothing in Semi-Supervised Graph Representation Learning

This report synthesises findings from 13 peer-reviewed papers addressing the following research question: What is the correlation between the number of message-passing layers and performance degradation in semi-supervised graph representation learning on Cora and Citeseer. Graph Neural Networks (GNNs) have achieved promising performance on a wide range of graph-based tasks. Despite their success, one severe limitation of GNNs is the over-smoothing issue (indistinguishable representations of nodes in different classes). 8 claims were extracted from source literature; 8 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 9.0/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: What is the correlation between the number of message-passing layers and performance degradation in semi-supervised graph representation learning on Cora and Citeseer?

Autonomous literature synthesis. Automated review score: 9.0/10. Full text and citation available at Assignee Research.