Papers › Break the Ceiling: Stronger Multi-scale Deep Graph Convolutional Networks

Break the Ceiling: Stronger Multi-scale Deep Graph Convolutional Networks

5 Jun 2019NeurIPS 2019 12arXiv:1906.02174archive 2025-07-28

Sitao Luan, Mingde Zhao, Xiao-Wen Chang, Doina Precup

Recently, neural network based approaches have achieved significant improvement for solving large, complex, graph-structured problems. However, their bottlenecks still need to be addressed, and the advantages of multi-scale information and deep architectures have not been sufficiently exploited. In this paper, we theoretically analyze how existing Graph Convolutional Networks (GCNs) have limited expressive power due to the constraint of the activation functions and their architectures. We generalize spectral graph convolution and deep GCN in block Krylov subspace forms and devise two architectures, both with the potential to be scaled deeper but each making use of the multi-scale information in different ways. We further show that the equivalence of these two architectures can be established under certain conditions. On several node classification tasks, with or without the help of validation, the two new architectures achieve better performance compared to many state-of-the-art methods.

PaperPDFConference PDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="1906.02174")

Code

Syntology Ran 1 of 9 code samples harvested from 1 repository linked to this paper; 8 have no recorded run. Of those that ran: 1 ran · honoured contract.

By repository: official repository: 9 samples from 1 repository, 1 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

PwnerHarry/Stronger_GCN officialmentioned in papermentioned on GitHubpytorchMIT 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.

Code Syntology ran Syntology

9 samples harvested; 1 ran; 1 honoured the contract we drafted; 8 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · honoured contract
8unverified

Licence: 0 of the 9 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from PwnerHarry/Stronger_GCN. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

parse_index_file PwnerHarry/Stronger_GCN/TF2/initialize_dataset.py official repository ran · honoured contract MIT (permissive) · 5c3fa9402a9405bc · report
dense2sparse PwnerHarry/Stronger_GCN/TF2/utils.py official repository unverified MIT (permissive) · 52e586b4eb952bec · report
load_data PwnerHarry/Stronger_GCN/TF2/initialize_dataset.py official repository unverified MIT (permissive) · 1d3501af271d574f · report
load_dataset PwnerHarry/Stronger_GCN/PyTorch/utils.py official repository unverified MIT (permissive) · 641da0a18c406c30 · report
normalize PwnerHarry/Stronger_GCN/TF2/initialize_dataset.py official repository unverified MIT (permissive) · abb8b57670b4a0e1 · report
sparse_matrix2tensor PwnerHarry/Stronger_GCN/TF2/utils.py official repository unverified MIT (permissive) · 31f819ee9d673402 · report
split_dataset PwnerHarry/Stronger_GCN/PyTorch/utils.py official repository unverified MIT (permissive) · dc4505e1479e9db1 · report
split_dataset PwnerHarry/Stronger_GCN/TF2/utils.py official repository unverified MIT (permissive) · 9d66a089d3b2d75a · report
train PwnerHarry/Stronger_GCN/PyTorch/utils.py official repository unverified MIT (permissive) · 03af7551a7b7d050 · report

Tasks

Node ClassificationNode Classification on Non-Homophilic (Heterophilic) Graphs

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Node Classification Chameleon (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 65.49 ± 1.64 #16 of 38 Archive leaderboard report
Node Classification Chameleon (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 64.99 ± 2.39 #17 of 38 Archive leaderboard report
Node Classification CiteSeer (0.5%) Truncated Krylov Accuracy 64.64% #3 of 14 Archive leaderboard report
Node Classification CiteSeer (0.5%) Snowball (tanh) Accuracy 62.05% #4 of 14 Archive leaderboard report
Node Classification CiteSeer (0.5%) Snowball (linear + tanh) Accuracy 61.99% #5 of 14 Archive leaderboard report
Node Classification CiteSeer (0.5%) Snowball (linear) Accuracy 59.41% #6 of 14 Archive leaderboard report
Node Classification CiteSeer (1%) Truncated Krylov Accuracy 69.03% #2 of 14 Archive leaderboard report
Node Classification CiteSeer (1%) Snowball (linear + tanh) Accuracy 67.07% #4 of 14 Archive leaderboard report
Node Classification CiteSeer (1%) Snowball (linear) Accuracy 65.85% #5 of 14 Archive leaderboard report
Node Classification CiteSeer (1%) Snowball (tanh) Accuracy 64.23% #6 of 14 Archive leaderboard report
Node Classification CiteSeer (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 81.53 ± 1.71 #15 of 33 Archive leaderboard report
Node Classification CiteSeer (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 80.93 ± 1.32 #19 of 33 Archive leaderboard report
Node Classification CiteSeer with Public Split: fixed 20 nodes per class Truncated Krylov Accuracy 73.86% #11 of 40 Archive leaderboard report
Node Classification CiteSeer with Public Split: fixed 20 nodes per class Snowball (tanh) Accuracy 73.32% #16 of 40 Archive leaderboard report
Node Classification CiteSeer with Public Split: fixed 20 nodes per class Snowball (linear) Accuracy 72.85% #22 of 40 Archive leaderboard report
Node Classification Cora (0.5%) Truncated Krylov Accuracy 74.89% #3 of 15 Archive leaderboard report
Node Classification Cora (0.5%) Snowball (tanh) Accuracy 71.36% #4 of 15 Archive leaderboard report
Node Classification Cora (0.5%) Snowball (linear) Accuracy 69.99% #5 of 15 Archive leaderboard report
Node Classification Cora (0.5%) Snowball (linear + tanh) Accuracy 67.76% #6 of 15 Archive leaderboard report
Node Classification Cora (1%) Truncated Krylov Accuracy 78.15% #2 of 15 Archive leaderboard report
Node Classification Cora (1%) Snowball (linear + tanh) Accuracy 74.79% #4 of 15 Archive leaderboard report
Node Classification Cora (1%) Snowball (tanh) Accuracy 74.78% #5 of 15 Archive leaderboard report
Node Classification Cora (1%) Snowball (linear) Accuracy 73.10% #6 of 15 Archive leaderboard report
Node Classification Cora (3%) Truncated Krylov Accuracy 81.92% #3 of 15 Archive leaderboard report
Node Classification Cora (3%) Snowball (linear) Accuracy 80.96% #4 of 15 Archive leaderboard report
Node Classification Cora (3%) Snowball (tanh) Accuracy 80.72% #5 of 15 Archive leaderboard report
Node Classification Cora (3%) Snowball (linear + tanh) Accuracy 79.52% #6 of 15 Archive leaderboard report
Node Classification Cora (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 89.33 ± 1.3 #7 of 33 Archive leaderboard report
Node Classification Cora (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 88.64 ± 1.15 #18 of 33 Archive leaderboard report
Node Classification Cora with Public Split: fixed 20 nodes per class Snowball (linear) Accuracy 83.26% #18 of 36 Archive leaderboard report
Node Classification Cora with Public Split: fixed 20 nodes per class Snowball (tanh) Accuracy 83.19% #20 of 36 Archive leaderboard report
Node Classification Cora with Public Split: fixed 20 nodes per class Truncated Krylov Accuracy 83.16% #21 of 36 Archive leaderboard report
Node Classification Cornell (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 82.95 ± 2.1 #23 of 36 Archive leaderboard report
Node Classification Cornell (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 82.62 ± 2.34 #24 of 36 Archive leaderboard report
Node Classification Film (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 36.00 ± 1.36 #27 of 37 Archive leaderboard report
Node Classification Film (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 35.97 ± 0.66 #29 of 37 Archive leaderboard report
Node Classification PubMed (0.03%) Truncated Krylov Accuracy 71.11% #2 of 14 Archive leaderboard report
Node Classification PubMed (0.03%) Snowball (linear) Accuracy 68.12% #3 of 14 Archive leaderboard report
Node Classification PubMed (0.03%) Snowball (tanh) Accuracy 62.61% #5 of 14 Archive leaderboard report
Node Classification PubMed (0.03%) Snowball (linear + tanh) Accuracy 61.94% #6 of 14 Archive leaderboard report
Node Classification PubMed (0.05%) Truncated Krylov Accuracy 72.57% #2 of 14 Archive leaderboard report
Node Classification PubMed (0.05%) Snowball (linear) Accuracy 70.04% #3 of 14 Archive leaderboard report
Node Classification PubMed (0.05%) Snowball (linear + tanh) Accuracy 69.45% #5 of 14 Archive leaderboard report
Node Classification PubMed (0.05%) Snowball (tanh) Accuracy 68.99% #6 of 14 Archive leaderboard report
Node Classification PubMed (0.1%) Truncated Krylov Accuracy 77.21% #1 of 14 Archive leaderboard report
Node Classification PubMed (0.1%) Snowball (linear + tanh) Accuracy 75.30% #3 of 14 Archive leaderboard report
Node Classification PubMed (0.1%) Snowball (tanh) Accuracy 74.40% #4 of 14 Archive leaderboard report
Node Classification PubMed (0.1%) Snowball (linear) Accuracy 73.83% #5 of 14 Archive leaderboard report
Node Classification PubMed (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 89.04 ± 0.49 #23 of 37 Archive leaderboard report
Node Classification PubMed (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 88.8 ± 0.82 #25 of 37 Archive leaderboard report
Node Classification PubMed with Public Split: fixed 20 nodes per class Truncated Krylov Accuracy 81.7% #6 of 37 Archive leaderboard report
Node Classification PubMed with Public Split: fixed 20 nodes per class Snowball (tanh) Accuracy 79.16% #21 of 37 Archive leaderboard report
Node Classification PubMed with Public Split: fixed 20 nodes per class Snowball (linear) Accuracy 79.10% #22 of 37 Archive leaderboard report
Node Classification Squirrel (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 48.25 ± 0.94 #16 of 37 Archive leaderboard report
Node Classification Squirrel (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 47.88 ± 1.23 #17 of 37 Archive leaderboard report
Node Classification Texas (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 83.11 ± 3.2 #25 of 36 Archive leaderboard report
Node Classification Texas (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 83.11 ± 3.2 #26 of 36 Archive leaderboard report
Node Classification Wisconsin (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 74.88 ± 3.42 #27 of 35 Archive leaderboard report
Node Classification Wisconsin (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 69.5 ± 5.01 #32 of 35 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Chameleon(60%/20%/20% random splits) Snowball-3 1:1 Accuracy 65.49 ± 1.64 #13 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Chameleon(60%/20%/20% random splits) Snowball-2 1:1 Accuracy 64.99 ± 2.39 #14 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Cornell (60%/20%/20% random splits) Snowball-3 1:1 Accuracy 82.95 ± 2.1 #23 of 33 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Cornell (60%/20%/20% random splits) Snowball-2 1:1 Accuracy 82.62 ± 2.34 #24 of 33 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Texas(60%/20%/20% random splits) Snowball-2 1:1 Accuracy 83.11 ± 3.2 #23 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Texas(60%/20%/20% random splits) Snowball-3 1:1 Accuracy 83.11 ± 3.2 #24 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Wisconsin(60%/20%/20% random splits) Snowball-2 1:1 Accuracy 74.88 ± 3.42 #24 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Wisconsin(60%/20%/20% random splits) Snowball-3 1:1 Accuracy 69.5 ± 5.01 #29 of 32 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.

Methods

ConvolutionGCNGraph Convolutional Networks

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