Accuracy on Cora (Node classification)
89.46AccuracyBES
Evaluation Results
| Method | Links | |
|---|---|---|
| BES2026.06 | 89.46 | |
| GraphECL2026.06 | 88.72 | |
| GCN2026.05 | 88.7 | |
| SAGEConv2026.05 | 88.7 | |
| H2GCN2026.06 | 88.69 | |
| IGCL2026.06 | 88.6 | |
| HGNN2026.05 | 88.5 | |
| SGC2026.05 | 88.4 | |
| SGC2026.06 | 88.29 | |
| fCWN2026.05 | 88.1 | |
| SSGC2026.06 | 87.86 | |
| GCN2026.06 | 87.74 | |
| GraphSAGE2026.06 | 87.74 | |
| SGNN2026.06 | 87.74 | |
| MotifRGC-GCNBackbone=GCN2026.06 | 87.73 | |
| MotifRGC-GATBackbone=GAT2026.06 | 87.62 | |
| MotifRGC-SAGEBackbone=GraphSAGE2026.06 | 87.54 | |
| GAT2026.05 | 87.5 | |
| KGNN2026.06 | 87.49 | |
| Cheb2026.06 | 87.31 | |
| sCWNc2026.05 | 87.3 | |
| GAT2026.06 | 87.18 | |
| GATv22026.06 | 86.32 | |
| GIN2026.05 | 86.2 | |
| MPNN-SAGEBackbone=GraphSAGE2026.06 | 86.2 | |
| MPNN-GATBackbone=GAT2026.06 | 86.14 | |
| MPNN-GCNBackbone=GCN2026.06 | 85.52 | |
| DiGGR2024.08 | 84.96 | |
| GiGaMAE2024.08 | 84.72 | |
| Bandana2024.08 | 84.62 | |
| GraphMAE22024.08 | 84.5 | |
| GraphMAEMethod Category=Feature-based2024.10 | 84.5 | |
| lrGAE 8Method Category=Structure-based2024.10 | 84.5 | |
| MaskGAEMethod Category=Structure-based2024.10 | 84.4 | |
| VEPM2024.08 | 84.3 | |
| SEEGERA2024.08 | 84.3 | |
| CCA-SSG2024.08 | 84.2 | |
| GraphMAE2024.08 | 84.2 | |
| lrGAE 7Method Category=Structure-based2024.10 | 84.2 | |
| IPGDN2024.08 | 84.1 | |
| SFABT2024.08 | 84.1 | |
| S2GAEMethod Category=Structure-based2024.10 | 84.1 | |
| SeeGeraMethod Category=Feature-based2024.10 | 84 | |
| lrGAE 6Method Category=Structure-based2024.10 | 84 | |
| AUG-MAEMethod Category=Feature-based2024.10 | 83.9 | |
| DisenGCN2024.08 | 83.7 | |
| S3GCLMethod Category=Standard GCL2024.10 | 83.7 | |
| GraphMAE2Method Category=Feature-based2024.10 | 83.7 | |
| EPAGCLMethod Category=Standard GCL2024.10 | 83.6 | |
| MVGRL2024.08 | 83.5 | |
| InfoGCL2024.08 | 83.5 | |
| CCA-SSGMethod Category=Standard GCL2024.10 | 83.5 | |
| MA-GCL2024.08 | 83.3 | |
| GGDMethod Category=Standard GCL2024.10 | 83.2 | |
| GiGaMAEMethod Category=Feature-based2024.10 | 83.2 | |
| GAT2024.08 | 83 | |
| GRACEMethod Category=Standard GCL2024.10 | 82.9 | |
| BGRLMethod Category=Standard GCL2024.10 | 82.8 | |
| BGRL2024.08 | 82.7 | |
| DGI2024.08 | 82.3 | |
| DGIMethod Category=Standard GCL2024.10 | 82.3 | |
| NodeImport-GCNBackbone=GCN2026.06 | 82.14 | |
| GRACE2024.08 | 81.9 | |
| GCN2024.08 | 81.5 | |
| SheafHyperGNN2026.05 | 81.3 | |
| SheafHyperGNNArchitecture Tier=Native Tier (set-function / message passing), Hidden dimension=64, Training Recipe=architecture-specific recipes2026.05 | 81.3 | |
| NodeImport-GATBackbone=GAT2026.06 | 81.18 | |
| CWN2026.05 | 81.1 | |
| RADE-OFBackbone=GCN2026.05 | 81.08 | |
| RADE-OFSBackbone=GCN2026.05 | 80.82 | |
| DropNodeBackbone=GCN2026.05 | 80.8 | |
| DropoutBackbone=GCN2026.05 | 80.76 | |
| DropMessageBackbone=GCN2026.05 | 80.65 | |
| ED-HNN2026.05 | 80.3 | |
| ED-HNNArchitecture Tier=Native Tier (set-function / message passing), Hidden dimension=64, Training Recipe=architecture-specific recipes2026.05 | 80.3 | |
| DropEdgeBackbone=GCN2026.05 | 80.28 | |
| GCNBackbone=GCN2026.05 | 80.1 | |
| NodeImport-SAGEBackbone=GraphSAGE2026.06 | 79.96 | |
| GAEMethod Category=Structure-based2024.10 | 79.8 | |
| SERBackbone=GT, Regularization Strategy=Structure Entropy Regularization2026.05 | 78.96 | |
| LapRBackbone=GT, Regularization Strategy=Laplacian Regularization2026.05 | 78.69 | |
| R-regBackbone=GT, Regularization Strategy=R-reg2026.05 | 78.57 | |
| DENSNET-D2026.05 | 77.5 | |
| DENSNET-DArchitecture Tier=Density-Aware Tier (ours), Hidden dimension=128, Training Recipe=shared ANCS recipe2026.05 | 77.5 | |
| CPBackbone=GT, Regularization Strategy=Canonical Polyadic2026.05 | 77.43 | |
| UniGNN2026.05 | 77.3 | |
| UniGNNArchitecture Tier=Native Tier (set-function / message passing), Hidden dimension=64, Training Recipe=shared ANCS recipe2026.05 | 77.3 | |
| VanillaBackbone=GT, Regularization Strategy=None2026.05 | 76.6 | |
| VGAE2024.08 | 76.3 | |
| AllDeepSets2026.05 | 76.2 | |
| AllDeepSetsArchitecture Tier=Native Tier (set-function / message passing), Hidden dimension=64, Training Recipe=shared ANCS recipe2026.05 | 76.2 | |
| P-regBackbone=GT, Regularization Strategy=P-reg2026.05 | 75.8 | |
| MLP2026.05 | 74.1 | |
| MLPArchitecture Tier=Feature-only reference, Hidden dimension=64, Training Recipe=shared ANCS recipe2026.05 | 74.1 | |
| HNHN2026.05 | 72.8 | |
| HNHNArchitecture Tier=Native Tier (set-function / message passing), Hidden dimension=64, Training Recipe=shared ANCS recipe2026.05 | 72.8 | |
| GAE2024.08 | 71.5 | |
| HyperGCN2026.05 | 71 | |
| HyperGCNArchitecture Tier=CE Tier (clique expansion), Hidden dimension=64, Training Recipe=shared ANCS recipe2026.05 | 71 | |
| HGNN2026.05 | 69.9 |