Graph Neural Networks struggle to capture long-range dependencies due to over-squashing, where information from exponentially growing neighborhoods must pass through a small number of structural bottlenecks. While recent rewiring methods attempt to alleviate this limitation, many rely on local criteria such as curvature, which can overlook global connectivity constraints that restrict information flow. We introduce Effective Resistance Rewiring (ERR), a simple topology correction strategy that uses effective resistance as a global signal to detect structural bottlenecks. ERR iteratively adds edges between node pairs with the largest resistance while removing edges with minimal resistance, strengthening weak communication pathways while controlling graph densification under a fixed edge budget. The procedure is parameter-free beyond the rewiring budget and relies on a single global measure aggregating all paths between node pairs. Beyond predictive performance with GCN models, we analyze how rewiring affects message propagation. By tracking cosine similarity between node embeddings across layers, we examine how the relationship between initial node features and learned representations evolves during message passing, comparing graphs with and without rewiring. This analysis helps determine whether improvements arise from better long-range communication rather than changes in embedding geometry. Experiments on homophilic and heterophilic graphs, including directed settings with DirGCN, reveal a trade-off between over-squashing and oversmoothing, where oversmoothing corresponds to the loss of representation diversity across layers. Resistance-guided rewiring improves connectivity and signal propagation but can accelerate representation mixing in deep models. Combining ERR with normalization techniques such as PairNorm stabilizes this trade-off and improves performance.
Mitigating Over-Squashing in Graph Neural Networks by Spectrum-Preserving Sparsification
Kim Jaechul Graduate School of Artificial Intelligence, KAIST, Seoul, Republic of Korea(韩国釜山科学技术院人工智能研究生院);School of Electrical Engineering, KAIST, Daejeon, Republic of Korea(韩国釜山科学技术院电子工程学院);Department of Engineering, University of Cambridge, Cambridge, United Kingdom(剑桥大学工程系);Incubation Institute, Fudan University, Shanghai, China(复旦大学孵化院);Shanghai Academy of Artificial Intelligence for Science, Shanghai, China(上海人工智能科学研究院);School of Information and Communication Technology, Griffith University, Gold Coast, Australia(格里菲斯大学信息与通信技术学院)
On the Trade-off between Over-smoothing and Over-squashing in Deep Graph Neural Networks
Télécom Paris - Institut Polytechnique de Paris(巴黎理工学院巴黎电信学院);BLUAI;La Rochelle Université(拉罗谢尔大学);Université Paris-Saclay(巴黎萨克雷大学);CentraleSupélec(中央理工-高等电力学院);Inria(法国国家信息与自动化研究所)
Revisiting Over-smoothing and Over-squashing Using Ollivier-Ricci Curvature
FPT Software AI Center(FPT软件人工智能中心);University of Science, Vietnam National University Ho Chi Minh City(越南国家大学胡志明市分校科学大学);University of Texas at Austin(德克萨斯大学奥斯汀分校);University of California, Los Angeles(加州大学洛杉矶分校);National University of Singapore(新加坡国立大学)
The Institute for Artificial Intelligence of Serbia(塞尔维亚人工智能研究所);McGill University(麦吉尔大学);Mila – Quebec AI Institute(米拉-魁北克人工智能研究所);Google DeepMind(谷歌DeepMind)
RAwR: Role-Aware Rewiring via Approximate Equitable Partition
Department of Excellence L’EMbeDS, Sant’Anna School of Advanced Studies, Pisa, Italy(卓越部门L’EMbeDS,圣安娜高级研究学院,意大利比萨);Department of Computer Science, University of Pisa, Italy(计算机科学系,意大利比萨大学);Department of Statistics, The Pennsylvania State University, USA(统计系,美国宾夕法尼亚州立大学);Huck Institutes of the Life Sciences, The Pennsylvania State University, USA(生命科学学院Huck研究所,美国宾夕法尼亚州立大学);IMT School for Advanced Studies, Lucca, Italy(IMT高级研究学院,意大利卢卡);Computational Network Science, RWTH Aachen University, Aachen, Germany(计算网络科学,德国亚琛工业大学,亚琛);DTU Technical University of Denmark, Lyngby, Denmark(丹麦技术大学DTU,Lyngby,丹麦)