Papers › DeeperGCN: All You Need to Train Deeper GCNs

DeeperGCN: All You Need to Train Deeper GCNs

13 Jun 2020arXiv:2006.07739archive 2025-07-28

Guohao Li, Chenxin Xiong, Ali Thabet, Bernard Ghanem

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.

PaperPDFConference PDFCode

In Syntology View this paper on Syntology, its page in Syntology's graph. That page lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

lightaime/deep_gcns_torch mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Training Graph Neural Networks with 1000 Layers (arXiv:2106.07476; that paper's own run record: 11 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 0 violated, 10 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
xnuohz/DeeperGCN-dgl mentioned on GitHubpytorchSyntology: not harvested report
dmlc/dgl pytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to node2vec: Scalable Feature Learning for Networks (arXiv:1607.00653; that paper's own run record: 15 ran (of which 2 constructed an object rather than computing a result; 12 with no instrument failure: 2 honoured, 0 violated, 10 with no contract checked; 3 where Syntology's instrument failed) · 10 unverified), Interaction Networks for Learning about Objects, Relations and Physics (arXiv:1612.00222; that paper's own run record: 6 ran (of which 6 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified; every one of the 6 samples that ran constructed an object rather than computing a result), Junction Tree Variational Autoencoder for Molecular Graph Generation (arXiv:1802.04364; that paper's own run record: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; the one sample that ran constructed an object rather than computing a result) and 3 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

AllGraph LearningGraph Property PredictionNode Property PredictionProperty PredictionRepresentation Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Graph Property Prediction ogbg-molhiv DeeperGCN Ext. data No #29 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv DeeperGCN Number of params 531976 #29 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv DeeperGCN Test ROC-AUC 0.7858 ± 0.0117 #29 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv DeeperGCN Validation ROC-AUC 0.8427 ± 0.0063 #29 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molpcba DeeperGCN+virtual node Ext. data No #27 of 36 Archive leaderboard report
Graph Property Prediction ogbg-molpcba DeeperGCN+virtual node Number of params 5550208 #27 of 36 Archive leaderboard report
Graph Property Prediction ogbg-molpcba DeeperGCN+virtual node Test AP 0.2781 ± 0.0038 #27 of 36 Archive leaderboard report
Graph Property Prediction ogbg-molpcba DeeperGCN+virtual node Validation AP 0.2920 ± 0.0025 #27 of 36 Archive leaderboard report
Graph Property Prediction ogbg-ppa DeeperGCN Ext. data No #10 of 18 Archive leaderboard report
Graph Property Prediction ogbg-ppa DeeperGCN Number of params 2336421 #10 of 18 Archive leaderboard report
Graph Property Prediction ogbg-ppa DeeperGCN Test Accuracy 0.7712 ± 0.0071 #10 of 18 Archive leaderboard report
Graph Property Prediction ogbg-ppa DeeperGCN Validation Accuracy 0.7313 ± 0.0078 #10 of 18 Archive leaderboard report
Link Property Prediction ogbl-collab DeeperGCN Ext. data No #22 of 34 Archive leaderboard report
Link Property Prediction ogbl-collab DeeperGCN Number of params 117383 #22 of 34 Archive leaderboard report
Link Property Prediction ogbl-collab DeeperGCN Test Hits@50 0.5273 ± 0.0047 #22 of 34 Archive leaderboard report
Link Property Prediction ogbl-collab DeeperGCN Validation Hits@50 0.6187 ± 0.0045 #22 of 34 Archive leaderboard report
Node Property Prediction ogbn-arxiv DeeperGCN Ext. data No #74 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv DeeperGCN Number of params 491176 #74 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv DeeperGCN Test Accuracy 0.7192 ± 0.0016 #74 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv DeeperGCN Validation Accuracy 0.7262 ± 0.0014 #74 of 86 Archive leaderboard report
Node Property Prediction ogbn-products DeeperGCN Ext. data No #40 of 64 Archive leaderboard report
Node Property Prediction ogbn-products DeeperGCN Number of params 253743 #40 of 64 Archive leaderboard report
Node Property Prediction ogbn-products DeeperGCN Test Accuracy 0.8098 ± 0.0020 #40 of 64 Archive leaderboard report
Node Property Prediction ogbn-products DeeperGCN Validation Accuracy 0.9238 ± 0.0009 #40 of 64 Archive leaderboard report
Node Property Prediction ogbn-proteins DeeperGCN Ext. data No #14 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins DeeperGCN Number of params 2374568 #14 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins DeeperGCN Test ROC-AUC 0.8580 ± 0.0017 #14 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins DeeperGCN Validation ROC-AUC 0.9106 ± 0.0016 #14 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins GEN + FLAG + node2vec Ext. data No #18 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins GEN + FLAG + node2vec Number of params 487436 #18 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins GEN + FLAG + node2vec Test ROC-AUC 0.8251 ± 0.0043 #18 of 26 Archive leaderboard report
Node Property Prediction ogbn-proteins GEN + FLAG + node2vec Validation ROC-AUC 0.8656 ± 0.0037 #18 of 26 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections