{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deepergcn-all-you-need-to-train-deeper-gcns","title":"DeeperGCN: All You Need to Train Deeper GCNs","arxiv_id":"2006.07739","date":"2020-06-13","proceeding":null,"authors":["Guohao Li","Chenxin Xiong","Ali Thabet","Bernard Ghanem"],"abstract":"Graph Convolutional Networks (GCNs) have been drawing significant attention with the power of representation learning on graphs. Unlike Convolutional Neural Networks (CNNs), which are able to take advantage of stacking very deep layers, GCNs suffer from vanishing gradient, over-smoothing and over-fitting issues when going deeper. These challenges limit the representation power of GCNs on large-scale graphs. This paper proposes DeeperGCN that is capable of successfully and reliably training very deep GCNs. We define differentiable generalized aggregation functions to unify different message aggregation operations (e.g. mean, max). We also propose a novel normalization layer namely MsgNorm and a pre-activation version of residual connections for GCNs. Extensive experiments on Open Graph Benchmark (OGB) show DeeperGCN significantly boosts performance over the state-of-the-art on the large scale graph learning tasks of node property prediction and graph property prediction. Please visit https://www.deepgcns.org for more information.","url_abs":"https://arxiv.org/abs/2006.07739v1","url_pdf":"https://arxiv.org/pdf/2006.07739v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deepergcn-all-you-need-to-train-deeper-gcns","repo_url":"https://github.com/lightaime/deep_gcns_torch/tree/master/examples/ogb_eff/ogbn_arxiv_dgl","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"deepergcn-all-you-need-to-train-deeper-gcns","repo_url":"https://github.com/xnuohz/DeeperGCN-dgl","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"deepergcn-all-you-need-to-train-deeper-gcns","repo_url":"https://github.com/dmlc/dgl/tree/master/examples/pytorch/deepergcn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"all","task_name":"All"},{"task_slug":"graph-learning","task_name":"Graph Learning"},{"task_slug":"graph-property-prediction","task_name":"Graph Property Prediction"},{"task_slug":"node-property-prediction","task_name":"Node Property Prediction"},{"task_slug":"property-prediction","task_name":"Property Prediction"},{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/graph-property-prediction-on-ogbg-molhiv","task":"Graph Property Prediction","dataset":"ogbg-molhiv","model":"DeeperGCN","rank_in_archive_order":29,"of":43,"metrics":{"Ext. data":"No","Number of params":"531976","Test ROC-AUC":"0.7858 ± 0.0117","Validation ROC-AUC":"0.8427 ± 0.0063"},"uses_additional_data":false},{"leaderboard":"/sota/graph-property-prediction-on-ogbg-molpcba","task":"Graph Property Prediction","dataset":"ogbg-molpcba","model":"DeeperGCN+virtual node","rank_in_archive_order":27,"of":36,"metrics":{"Ext. data":"No","Number of params":"5550208","Test AP":"0.2781 ± 0.0038","Validation AP":"0.2920 ± 0.0025"},"uses_additional_data":false},{"leaderboard":"/sota/graph-property-prediction-on-ogbg-ppa","task":"Graph Property Prediction","dataset":"ogbg-ppa","model":"DeeperGCN","rank_in_archive_order":10,"of":18,"metrics":{"Ext. data":"No","Number of params":"2336421","Test Accuracy":"0.7712 ± 0.0071","Validation Accuracy":"0.7313 ± 0.0078"},"uses_additional_data":false},{"leaderboard":"/sota/link-property-prediction-on-ogbl-collab","task":"Link Property Prediction","dataset":"ogbl-collab","model":"DeeperGCN","rank_in_archive_order":22,"of":34,"metrics":{"Ext. data":"No","Number of params":"117383","Test Hits@50":"0.5273 ± 0.0047","Validation Hits@50":"0.6187 ± 0.0045"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-arxiv","task":"Node Property Prediction","dataset":"ogbn-arxiv","model":"DeeperGCN","rank_in_archive_order":74,"of":86,"metrics":{"Ext. data":"No","Number of params":"491176","Test Accuracy":"0.7192 ± 0.0016","Validation Accuracy":"0.7262 ± 0.0014"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-products","task":"Node Property Prediction","dataset":"ogbn-products","model":"DeeperGCN","rank_in_archive_order":40,"of":64,"metrics":{"Ext. data":"No","Number of params":"253743","Test Accuracy":"0.8098 ± 0.0020","Validation Accuracy":"0.9238 ± 0.0009"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-proteins","task":"Node Property Prediction","dataset":"ogbn-proteins","model":"DeeperGCN","rank_in_archive_order":14,"of":26,"metrics":{"Ext. data":"No","Number of params":"2374568","Test ROC-AUC":"0.8580 ± 0.0017","Validation ROC-AUC":"0.9106 ± 0.0016"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-proteins","task":"Node Property Prediction","dataset":"ogbn-proteins","model":"GEN + FLAG + node2vec","rank_in_archive_order":18,"of":26,"metrics":{"Ext. data":"No","Number of params":"487436","Test ROC-AUC":"0.8251 ± 0.0043","Validation ROC-AUC":"0.8656 ± 0.0037"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2006.07739","atlas_url":"https://app.syntology.ai/?focus=2006.07739","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}