Node Classification on Coauthor CS
96.5AccuracySAGE†
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SAGE†Recipe=Deep2026.05 | 96.5 | — | |
| GraphSAGEImplementation=Ours2024.06 | 96.38 | — | |
| GAT†Recipe=Deep2026.05 | 96.25 | — | |
| GATImplementation=Ours2024.06 | 96.21 | — | |
| GCNImplementation=Literature2024.06 | 96.18 | — | |
| GCNImplementation=Ours2024.06 | 96.17 | — | |
| GCN†Recipe=Deep2026.05 | 96.13 | — | |
| AGDN-HC2026.06 | 96.04 | — | |
| OURSPipeline=A2026.05 | 95.99 | — | |
| ExphormerImplementation=Ours2024.06 | 95.92 | — | |
| ChebNetII2026.06 | 95.87 | — | |
| NAGphormerImplementation=Ours2024.06 | 95.85 | — | |
| AGDN-HA2026.06 | 95.81 | — | |
| NAGphormerImplementation=Literature2024.06 | 95.75 | — | |
| SGFormerImplementation=Ours2024.06 | 95.71 | — | |
| NodeFormerImplementation=Ours2024.06 | 95.69 | — | |
| ACMGCN2026.06 | 95.67 | — | |
| NodeFormerImplementation=Literature2024.06 | 95.64 | — | |
| FilterMoE2026.06 | 95.61 | — | |
| PolynormerImplementation=Literature2024.06 | 95.53 | — | |
| Polynormer2026.05 | 95.53 | — | |
| GCNJumping Knowledge (JK) Strategy=Pool2019.04 | 95.47 | — | |
| ADC2026.06 | 95.45 | — | |
| GCNJumping Knowledge (JK) Strategy=Concat2019.04 | 95.44 | — | |
| PolynormerImplementation=Ours2024.06 | 95.42 | — | |
| NodeMoE2026.06 | 95.38 | — | |
| Mowst2026.06 | 95.35 | — | |
| GPRGNN2026.06 | 95.13 | — | |
| GATJumping Knowledge (JK) Strategy=Concat2019.04 | 95.12 | — | |
| JacobiConv2026.06 | 95.06 | — | |
| ExphormerImplementation=Literature2024.06 | 94.93 | — | |
| GATJumping Knowledge (JK) Strategy=Pool2019.04 | 94.84 | — | |
| GCN-LPA2020.02 | 94.8 | — | |
| SGFormerImplementation=Literature2024.06 | 94.78 | — | |
| SGFormer2026.05 | 94.78 | — | |
| DSF2026.06 | 94.78 | — | |
| GAMLP2026.06 | 94.78 | — | |
| NFGNN2026.06 | 94.72 | — | |
| DNANumber of groups (g)=162019.04 | 94.64 | — | |
| HOGA2026.06 | 94.62 | — | |
| DNANumber of groups (g)=82019.04 | 94.46 | — | |
| GMoE2026.06 | 94.45 | — | |
| GCNJumping Knowledge (JK) Strategy=LSTM2019.04 | 94.4 | — | |
| GOATImplementation=Literature2024.06 | 94.21 | — | |
| GATJumping Knowledge (JK) Strategy=LSTM2019.04 | 94.09 | — | |
| GraphGPSImplementation=Ours2024.06 | 94.04 | — | |
| DNANumber of groups (g)=12019.04 | 94.02 | — | |
| GraphGPSImplementation=Literature2024.06 | 93.93 | — | |
| GraphGPS2026.05 | 93.93 | — | |
| GraphSAGEImplementation=Literature2024.06 | 93.91 | — | |
| GraphSAGE2026.06 | 93.91 | — | |
| SIGN2026.06 | 93.89 | — | |
| GOATImplementation=Ours2024.06 | 93.81 | — | |
| GAT2020.02 | 93.8 | — | |
| GATImplementation=Literature2024.06 | 93.61 | — | |
| GAT2026.06 | 93.61 | — | |
| GCN2020.02 | 93.6 | — | |
| GATJumping Knowledge (JK) Strategy=None2019.04 | 93.54 | — | |
| GNSN2026.05 | 93.31 | — | |
| GCN2023.06 | 93.03 | — | |
| GCNdiffusion=PPR2019.10 | 93.01 | — | |
| FROND2026.05 | 93 | — | |
| ACMP2026.05 | 93 | — | |
| GCN2026.06 | 92.92 | — | |
| GCNJumping Knowledge (JK) Strategy=None2019.04 | 92.9 | — | |
| GRAND-lvariant=l2024.03 | 92.9 | — | |
| GRAND2026.05 | 92.9 | — | |
| BLEND2026.05 | 92.9 | — | |
| GCNdiffusion=Heat2019.10 | 92.79 | — | |
| GCNIIMethod Category=Deep GNNs, Setting=Inductive2023.05 | 92.7 | — | |
| GRADE-Gaussian2024.03 | 92.7 | — | |
| ARMAdiffusion=PPR2019.10 | 92.63 | — | |
| GraphSAGEaggregator=mean2020.02 | 92.6 | — | |
| GEMMethod Category=Deep GNNs, Setting=Inductive2023.05 | 92.58 | — | |
| JKdiffusion=PPR2019.10 | 92.41 | — | |
| JK-Netaggregator=concat2020.02 | 92.4 | — | |
| JKdiffusion=Heat2019.10 | 92.4 | — | |
| ARMAdiffusion=Heat2019.10 | 92.32 | — | |
| GAT2023.06 | 92.31 | — | |
| CGNN2024.03 | 92.3 | — | |
| CGNN2026.05 | 92.3 | — | |
| GraphMLPMethod Category=MLP-based, Setting=Inductive2023.05 | 92.29 | — | |
| GCNdiffusion=AdaDIF2019.10 | 92.28 | — | |
| GLNNMethod Category=MLP-based, Setting=Inductive2023.05 | 92.27 | — | |
| DGI2023.06 | 92.15 | — | |
| OKDEEM0Method Category=MLP-based, Setting=Inductive, Inference Mode=Non-hop2023.05 | 92.13 | — | |
| GRADE-GAT2024.03 | 92.1 | — | |
| APPNPdiffusion=No diffusion2019.10 | 92.08 | — | |
| GCNdiffusion=No diffusion2019.10 | 91.83 | — | |
| GCNModel=GCN, Training Source=Target dataset2026.04 | 91.83 | — | |
| OKDEEMMethod Category=Stochastic GNNs, Setting=Inductive, Inference Mode=One-hop2023.05 | 91.81 | — | |
| ACMP-GAT2024.03 | 91.8 | — | |
| MLPhidden_layer_sizes=502020.02 | 91.7 | — | |
| JKdiffusion=AdaDIF2019.10 | 91.68 | — | |
| GDE2024.03 | 91.6 | — | |
| GDE2026.05 | 91.6 | — | |
| NodePFNTraining=Zero-shot from synthetic prior data2026.04 | 91.55 | — | |
| NodePFNModel=NodePFN, Training Source=Single pre-trained model2026.04 | 91.55 | — | |
| GCNNumber of runs=1002022.04 | 91.5 | — | |
| GPRGNNTraining=Dataset-specific2026.04 | 91.49 | — |