Summary
Over-squashing is a phenomenon in Graph neural networks where individual nodes are not sensitive to distant information in ways that cannot be ameliorated by increasing network depth (Di (1)). The term was coined by (2).
1.
Giovanni FD, Giusti L, Barbero F, Luise G, Lio P, Bronstein MM. On Over-Squashing in Message Passing Neural Networks: The Impact of Width, Depth, and Topology. In: International Conference on Machine Learning. PMLR; 2023. p. 7865–85. Available from: https://proceedings.mlr.press/v202/di-giovanni23a.html
2.
Alon U, Yahav E. On the Bottleneck of Graph Neural Networks and its Practical Implications. In: ICLR 2021. 2021. Available from: https://openreview.net/forum?id=i80OPhOCVH2