# GResNet: Graph Residual Network for Reviving Deep GNNs from Suspended Animation

arXiv: https://arxiv.org/abs/1909.05729 · HTML: https://mlsanity.com/paper/1909.05729/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Sep 2019.

## Built on by (most cited first, 22 of 22)

- [Transformer for Graphs: An Overview from Architecture Perspective](https://mlsanity.com/paper/2202.08455/) (2022)
- [Graph-ToolFormer: To Empower LLMs with Graph Reasoning Ability via Prompt Augmented by ChatGPT](https://mlsanity.com/paper/2304.11116/) (2023)
- [Dissecting the Diffusion Process in Linear Graph Convolutional Networks](https://mlsanity.com/paper/2102.10739/) (2021)
- [Learning Attributed Graph Representations with Communicative Message Passing Transformer](https://mlsanity.com/paper/2107.08773/) (2021)
- [How Expressive are Transformers in Spectral Domain for Graphs?](https://mlsanity.com/paper/2201.09332/) (2022)
- [Graph Neural Networks for Text Classification: A Survey](https://mlsanity.com/paper/2304.11534/) (2023)
- [Graph Convolutional Networks for Model-Based Learning in Nonlinear Inverse Problems](https://mlsanity.com/paper/2103.15138/) (2021)
- [Multi-Level Attention Pooling for Graph Neural Networks: Unifying Graph Representations with Multiple Localities](https://mlsanity.com/paper/2103.01488/) (2021)
- [Molecular Joint Representation Learning via Multi-modal Information](https://mlsanity.com/paper/2211.14042/) (2022)
- [Contrastive Representation Learning Based on Multiple Node-centered Subgraphs](https://mlsanity.com/paper/2308.16441/) (2023)
- [Graph network for learning bi-directional physics](https://mlsanity.com/paper/2112.07054/) (2021)
- [Multimodal Graph-based Transformer Framework for Biomedical Relation Extraction](https://mlsanity.com/paper/2107.00596/) (2021)
- [Measuring and Sampling: A Metric-guided Subgraph Learning Framework for Graph Neural Network](https://mlsanity.com/paper/2112.15015/) (2021)
- [A Bird's-Eye Tutorial of Graph Attention Architectures](https://mlsanity.com/paper/2206.02849/) (2022)
- [NaroNet: Discovery of tumor microenvironment elements from highly multiplexed images](https://mlsanity.com/paper/2103.05385/) (2021)
- [Multi-hop Graph Convolutional Network with High-order Chebyshev Approximation for Text Reasoning](https://mlsanity.com/paper/2106.05221/) (2021)
- [NDGGNET-A Node Independent Gate based Graph Neural Networks](https://mlsanity.com/paper/2205.05348/) (2022)
- [Iso-CapsNet: Isomorphic Capsule Network for Brain Graph Representation Learning](https://mlsanity.com/paper/2206.13465/) (2022)
- [RPN 2: On Interdependence Function Learning Towards Unifying and Advancing CNN, RNN, GNN, and Transformer](https://mlsanity.com/paper/2411.11162/) (2024)
- [Flex-GAD : Flexible Graph Anomaly Detection](https://mlsanity.com/paper/2510.25809/) (2025)
- [DynBERG: Dynamic BERT-based Graph neural network for financial fraud detection](https://mlsanity.com/paper/2511.00047/) (2025)
- [Shared-Roadmap Generation and Evaluator for Multi-Agent Path Planning Using Heterogeneous Graph Neural Network](https://mlsanity.com/paper/2610.09034/) (2026)

## Abstract (arXiv)

The existing graph neural networks (GNNs) based on the spectral graph convolutional operator have been criticized for its performance degradation, which is especially common for the models with deep architectures. In this paper, we further identify the suspended animation problem with the existing GNNs. Such a problem happens when the model depth reaches the suspended animation limit, and the model will not respond to the training data any more and become not learnable. Analysis about the causes of the suspended animation problem with existing GNNs will be provided in this paper, whereas several other peripheral factors that will impact the problem will be reported as well. To resolve the problem, we introduce the GResNet (Graph Residual Network) framework in this paper, which creates extensively connected highways to involve nodes' raw features or intermediate representations throughout the graph for all the model layers. Different from the other learning settings, the extensive connections in the graph data will render the existing simple residual learning methods fail to work. We prove the effectiveness of the introduced new graph residual terms from the norm preservation perspective, which will help avoid dramatic changes to the node's representations between sequential layers. Detailed studies about the GResNet framework for many existing GNNs, including GCN, GAT and LoopyNet, will be reported in the paper with extensive empirical experiments on real-world benchmark datasets.