Papers › Do We Need Anisotropic Graph Neural Networks?

Do We Need Anisotropic Graph Neural Networks?

3 Apr 2021arXiv:2104.01481archive 2025-07-28

Shyam A. Tailor, Felix L. Opolka, Pietro Liò, Nicholas D. Lane

Common wisdom in the graph neural network (GNN) community dictates that anisotropic models -- in which messages sent between nodes are a function of both the source and target node -- are required to achieve state-of-the-art performance. Benchmarks to date have demonstrated that these models perform better than comparable isotropic models -- where messages are a function of the source node only. In this work we provide empirical evidence challenging this narrative: we propose an isotropic GNN, which we call Efficient Graph Convolution (EGC), that consistently outperforms comparable anisotropic models, including the popular GAT or PNA architectures by using spatially-varying adaptive filters. In addition to raising important questions for the GNN community, our work has significant real-world implications for efficiency. EGC achieves higher model accuracy, with lower memory consumption and latency, along with characteristics suited to accelerator implementation, while being a drop-in replacement for existing architectures. As an isotropic model, it requires memory proportional to the number of vertices in the graph (𝒪(V)); in contrast, anisotropic models require memory proportional to the number of edges (𝒪(E)). We demonstrate that EGC outperforms existing approaches across 6 large and diverse benchmark datasets, and conclude by discussing questions that our work raise for the community going forward. Code and pretrained models for our experiments are provided at https://github.com/shyam196/egc.

PaperPDFCode

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

pyg-team/pytorch_geometric officialmentioned in papermentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Learning the Geometry of Data: A Mathematical Review of Shape Space Analysis (arXiv:2606.17022; that paper's own run record: 8 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 2 honoured, 0 violated, 4 with no contract checked; 2 where Syntology's instrument failed) · 4 unverified); nothing checks that they implement this paper's method. report
shyam196/egc officialmentioned in papermentioned on GitHubpytorchSyntology: not harvested 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

Graph Neural Network

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Graph Property Prediction ogbg-code2 EGC-M (No Edge Features) Ext. data No #12 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-M (No Edge Features) Number of params 10986002 #12 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-M (No Edge Features) Test F1 score 0.1595 ± 0.0019 #12 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-M (No Edge Features) Validation F1 score 0.1464 ± 0.0021 #12 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 PNA (No Edge Features) Ext. data No #15 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 PNA (No Edge Features) Number of params 10992050 #15 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 PNA (No Edge Features) Test F1 score 0.1570 ± 0.0032 #15 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 PNA (No Edge Features) Validation F1 score 0.1453 ± 0.0025 #15 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 MPNN-Max (No Edge Features) Ext. data No #17 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 MPNN-Max (No Edge Features) Number of params 10971506 #17 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 MPNN-Max (No Edge Features) Test F1 score 0.1552 ± 0.0022 #17 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 MPNN-Max (No Edge Features) Validation F1 score 0.1441 ± 0.0016 #17 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-S (No Edge Features) Ext. data No #18 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-S (No Edge Features) Number of params 11156530 #18 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-S (No Edge Features) Test F1 score 0.1528 ± 0.0025 #18 of 21 Archive leaderboard report
Graph Property Prediction ogbg-code2 EGC-S (No Edge Features) Validation F1 score 0.1427 ± 0.0020 #18 of 21 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-M (No Edge Features) Ext. data No #32 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-M (No Edge Features) Number of params 317265 #32 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-M (No Edge Features) Test ROC-AUC 0.7818 ± 0.0153 #32 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-M (No Edge Features) Validation ROC-AUC 0.8396 ± 0.0097 #32 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-S (No Edge Features) Ext. data No #36 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-S (No Edge Features) Number of params 317013 #36 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-S (No Edge Features) Test ROC-AUC 0.7721 ± 0.0110 #36 of 43 Archive leaderboard report
Graph Property Prediction ogbg-molhiv EGC-S (No Edge Features) Validation ROC-AUC 0.8366 ± 0.0074 #36 of 43 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-S (100k) Ext. data No #65 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-S (100k) Number of params 100648 #65 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-S (100k) Test Accuracy 0.7219 ± 0.0016 #65 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-S (100k) Validation Accuracy 0.7338 ± 0.0022 #65 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-M (100k) Ext. data No #71 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-M (100k) Number of params 99464 #71 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-M (100k) Test Accuracy 0.7196 ± 0.0023 #71 of 86 Archive leaderboard report
Node Property Prediction ogbn-arxiv EGC-M (100k) Validation Accuracy 0.7334 ± 0.0013 #71 of 86 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

ConvolutionGATGraph Neural NetworkPNA

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