Node Classification on Cornell (test)
91.8Mean AccuracyAPPNP
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| APPNPModel Type=Deep2022.02 | 91.8 | — | — | — | |
| MLPModel Type=Deep2022.02 | 91.36 | — | — | — | |
| GPRGNNModel Type=Deep2022.02 | 91.36 | — | — | — | |
| Node2Vec2026.06 | 90.5 | — | — | — | |
| DeepWalk2026.06 | 89.2 | — | — | — | |
| ASPECTEvaluation Protocol=linear protocol2026.04 | 88.85 | — | — | — | |
| G2-GATgradient_gating=true2022.10 | 87.3 | — | — | — | |
| FGN2026.06 | 87.1 | — | — | — | |
| GCN-IED2026.06 | 86.8 | — | — | — | |
| BundlePolySD2025.11 | 86.76 | — | — | — | |
| RawModel Type=Non-Deep2022.02 | 86.49 | — | — | — | |
| G2-GCNgradient_gating=true2022.10 | 86.49 | — | — | — | |
| DiagPolySD2025.11 | 86.49 | — | — | — | |
| GeneralPolySD2025.11 | 86.49 | — | — | — | |
| NSDrestriction maps=diagonal2025.11 | 86.49 | — | — | — | |
| ASGCModel Type=Non-Deep2022.02 | 86.22 | — | — | — | |
| G2-GraphSAGEgradient_gating=true2022.10 | 86.22 | — | — | — | |
| RiSNN2025.11 | 85.95 | — | — | — | |
| Conn-NSD2025.11 | 85.95 | — | — | — | |
| GGCN2022.10 | 85.68 | — | — | — | |
| GGCNHalf-Hop augmentation=false2023.08 | 85.68 | — | — | — | |
| NSDrestriction maps=general2025.11 | 85.68 | — | — | — | |
| GGCN2025.11 | 85.68 | — | — | — | |
| JdSNN2025.11 | 85.41 | — | — | — | |
| GCN-ChebyModel Type=Deep2022.02 | 85.33 | — | — | — | |
| MAGCN2026.06 | 85.3 | — | — | — | |
| GInterNet2026.06 | 85.3 | — | — | — | |
| EIGNNAttack=FGSM, Perturbation=0.00012022.02 | 85.13 | — | — | — | |
| EIGNNAttack=PGD, Perturbation=0.00012022.02 | 85.13 | — | — | — | |
| NLMLP2020.05 | 84.9 | — | — | — | |
| NSDrestriction maps=orthogonal2025.11 | 84.86 | — | — | — | |
| JdSNNweighting=none2025.11 | 84.59 | — | — | — | |
| ANSD2025.11 | 84.59 | — | — | — | |
| UDGNNBackbone=GCN2022.05 | 84.32 | — | — | — | |
| SAN2025.11 | 84.32 | — | — | — | |
| SimP-GCN2020.05 | 84.1 | — | — | — | |
| EIGNNAttack=FGSM, Perturbation=0.0012022.02 | 84.05 | — | — | — | |
| EIGNNAttack=PGD, Perturbation=0.0012022.02 | 84.05 | — | — | — | |
| F2GAT2022.10 | 83.51 | — | — | — | |
| GloGNN2022.10 | 83.51 | — | — | — | |
| GloGNN2026.06 | 83.5 | — | — | — | |
| UDGNNBackbone=SAGE2022.05 | 83.24 | — | — | — | |
| H2GCNAttack=PGD, Perturbation=0.00012022.02 | 82.97 | — | — | — | |
| RiSNNtransport=none2025.11 | 82.97 | — | — | — | |
| UDGNNBackbone=GAT2022.05 | 82.71 | — | — | — | |
| H2GCN2022.10 | 82.7 | — | — | — | |
| H2GCNHalf-Hop augmentation=false2023.08 | 82.7 | — | — | — | |
| H2GCN2025.11 | 82.7 | — | — | — | |
| POLYGCLEvaluation Protocol=linear protocol2026.04 | 82.62 | — | — | — | |
| UDGNN*Backbone=GCN2022.05 | 82.43 | — | — | — | |
| GPRGNNGraph modification variant=GDC2022.06 | 82.4 | — | — | — | |
| PPGNNGraph modification variant=Original2022.06 | 82.4 | — | — | — | |
| H2GCN2020.05 | 82.2 | — | — | — | |
| H2GCNGraph modification variant=Original2022.06 | 82.2 | — | — | — | |
| H2GCNbest model variant=true2021.06 | 82.16 | — | — | — | |
| H2GCN-12022.05 | 82.16 | — | — | — | |
| H2GCN-22022.05 | 82.16 | — | — | — | |
| UDGNN*Backbone=SAGE2022.05 | 82.16 | — | — | — | |
| GATGraph modification variant=Adaptive Spectral Clustering2022.06 | 81.9 | — | — | — | |
| UDGNN*Backbone=GAT2022.05 | 81.89 | — | — | — | |
| MLP2022.10 | 81.89 | — | — | — | |
| MLPHalf-Hop augmentation=false2023.08 | 81.89 | — | — | — | |
| MLP2025.11 | 81.89 | — | — | — | |
| WRGAT2021.06 | 81.62 | — | — | — | |
| MLP2020.05 | 81.6 | — | — | — | |
| APPNPGraph modification variant=Adaptive Spectral Clustering2022.06 | 81.6 | — | — | — | |
| S3GCLEvaluation Protocol=linear protocol2026.04 | 81.27 | — | — | — | |
| CHEVGraph modification variant=GDC2022.06 | 81.1 | — | — | — | |
| ARMAGraph modification variant=Adaptive Spectral Clustering2022.06 | 81.1 | — | — | — | |
| MLP2022.05 | 81.08 | — | — | — | |
| GPRGNN2022.05 | 81.08 | — | — | — | |
| CHEVGraph modification variant=Adaptive Spectral Clustering2022.06 | 80.8 | — | — | — | |
| FTCP2026.06 | 80.8 | — | — | — | |
| ELU-GCN2026.06 | 80.4 | — | — | — | |
| GGDEvaluation Protocol=linear protocol2026.04 | 80.33 | — | — | — | |
| GPRGNNGraph modification variant=Adaptive Spectral Clustering2022.06 | 80.3 | — | — | — | |
| BernNetGraph modification variant=Original2022.06 | 80.3 | — | — | — | |
| GPRGNN2022.10 | 80.27 | — | — | — | |
| GPRGNN2025.12 | 80.27 | — | — | — | |
| GPRGNN2025.11 | 80.27 | — | — | — | |
| GPRGNNGraph modification variant=Original2022.06 | 79.5 | — | — | — | |
| HGRN2026.06 | 79.5 | — | — | — | |
| H2GCNAttack=FGSM, Perturbation=0.00012022.02 | 79.46 | — | — | — | |
| PAGCN2026.06 | 79.3 | — | — | — | |
| GCNGraph modification variant=Adaptive Spectral Clustering2022.06 | 79.2 | — | — | — | |
| GCNII2022.05 | 79.19 | — | — | — | |
| FAGCN2022.10 | 79.19 | — | — | — | |
| FAGCN2025.11 | 79.19 | — | — | — | |
| DSHN2025.10 | 79.19 | — | — | 4.37 | |
| DSHNLight2025.10 | 79.19 | — | — | 3.2 | |
| H2GCN+LCC2026.06 | 78.92 | — | — | — | |
| CCA-SSGEvaluation Protocol=linear protocol2026.04 | 78.69 | — | — | — | |
| MOGCN2026.06 | 78.5 | — | — | — | |
| ARMAGraph modification variant=GDC2022.06 | 78.4 | — | — | — | |
| GeDi-HNN2025.10 | 78.37 | — | — | 3.19 | |
| GREETEvaluation Protocol=linear protocol2026.04 | 78.36 | — | — | — | |
| LoGoGNN2026.06 | 78.2 | — | — | — | |
| BAGCN2026.06 | 77.9 | — | — | — | |
| GCNII2022.10 | 77.86 | — | — | — | |
| GCNII2025.11 | 77.86 | — | — | — |