Graph Classification on COLLAB
88AccuracyGraphMAE
Evaluation Results
| Method | Links | |
|---|---|---|
| GraphMAEEvaluation protocol=Unsupervised representation learning2024.05 | 88 | |
| AD-GCLEvaluation protocol=Unsupervised representation learning2024.05 | 85.5 | |
| MVGRLEvaluation protocol=Unsupervised representation learning2024.05 | 84.5 | |
| CORE-WL-VH2024.06 | 84.5 | |
| CORE-WL-VHMethod Category=Kernels2025.10 | 84.5 | |
| GraphMAE2 + COREmasking process=GraphMAE22025.12 | 84.3 | |
| GraphMAE22025.12 | 84.1 | |
| PXGL-GNN2025.12 | 83.96 | |
| DiGGRLearning Paradigm=Unsupervised2024.08 | 83.76 | |
| GNTKinput=constant (X=1)2019.05 | 83.6 | |
| GraphMAE + COREmasking process=GraphMAE2025.12 | 83.4 | |
| AIA2022.11 | 82.98 | |
| M-Mixup2022.11 | 82.9 | |
| GCBM-E2026.03 | 82.8 | |
| SAN2025.12 | 82.73 | |
| GraphMAEApproach Category=Feature-based2024.10 | 82.7 | |
| CAL2022.11 | 82.68 | |
| FLAG2022.11 | 82.48 | |
| GraphCrocEpochs=1002024.10 | 82.4 | |
| GCBM2026.03 | 82.4 | |
| CDLLabel=70%2024.11 | 82.36 | |
| S2GAE2025.12 | 82.35 | |
| lrGAE 6Approach Category=Structure-based2024.10 | 82.3 | |
| lrGAE 7Approach Category=Structure-based2024.10 | 82.3 | |
| S2GAEApproach Category=Structure-based2024.10 | 82.2 | |
| lrGAE 8Approach Category=Structure-based2024.10 | 82.2 | |
| DiffPool-DETlearning_type=Inductive learning2019.05 | 82.13 | |
| ERM2022.11 | 82.08 | |
| MaskGAEApproach Category=Structure-based2024.10 | 82 | |
| Graphormer2026.07 | 81.8 | |
| GCNnode features=augmented with degree, input=constant (X=1)2019.05 | 81.72 | |
| DDMProtocol=Unsupervised representation learning2023.06 | 81.72 | |
| GraphCrocEpochs=102024.10 | 81.7 | |
| GraphMAE2Approach Category=Feature-based2024.10 | 81.7 | |
| GFNK=3, node features=degree + multi-scale, input=constant (X=1)2019.05 | 81.5 | |
| GFNInference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 81.5 | |
| GAWL2026.03 | 81.5 | |
| GFN# Parameters=68,7542023.11 | 81.5 | |
| GFN2023.11 | 81.5 | |
| DisenSemiLearning Paradigm=Supervised/Semi-supervised2024.08 | 81.5 | |
| BGRLApproach Category=Standard GCL2024.10 | 81.5 | |
| ICL2025.12 | 81.45 | |
| WWL2026.03 | 81.4 | |
| GraphGPS2026.07 | 81.4 | |
| PPGNInference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 81.38 | |
| GFN-lightK=3, node features=degree + multi-scale, input=constant (X=1)2019.05 | 81.34 | |
| GCMAE2025.12 | 81.3 | |
| SEP2026.03 | 81.3 | |
| SEP-G2024.10 | 81.28 | |
| SubGNN2025.12 | 81.26 | |
| GIC2018.11 | 81.24 | |
| DGCLLearning Paradigm=Unsupervised2024.08 | 81.2 | |
| GLIALabel=70%2024.11 | 81.18 | |
| NH2024.06 | 81.1 | |
| NHMethod Category=Kernels2025.10 | 81.1 | |
| S2GAE2024.10 | 81.02 | |
| RetGK_Iinput=constant (X=1)2019.05 | 81 | |
| RetGk2018.11 | 81 | |
| CDLLabel=50%2024.11 | 80.96 | |
| CaCoSE2026.03 | 80.95 | |
| MINCUTPOOL2026.03 | 80.9 | |
| (s)GCNneighborhood enlargement=true2019.05 | 80.89 | |
| MinCutPool2024.10 | 80.87 | |
| GCLLabel=70%2024.11 | 80.82 | |
| SeeGeraApproach Category=Feature-based2024.10 | 80.8 | |
| Trans. with X-LogSMaskvariant=ours2026.07 | 80.8 | |
| GMT2024.10 | 80.74 | |
| GMT2026.07 | 80.74 | |
| (s)GIN-0neighborhood enlargement=true, epsilon=02019.05 | 80.71 | |
| WL-OAKernel Type=Weisfeiler-Lehman Optimal Assignment2016.06 | 80.7 | |
| WL-OA2018.11 | 80.7 | |
| OT-GNN2026.03 | 80.7 | |
| GMT2026.03 | 80.7 | |
| S3GCLApproach Category=Standard GCL2024.10 | 80.7 | |
| GatedGCN2026.07 | 80.7 | |
| DYF2018.11 | 80.61 | |
| GCNPooling Strategy=Flat2019.03 | 80.6 | |
| RetGK_IIinput=constant (X=1)2019.05 | 80.6 | |
| SAT2026.03 | 80.6 | |
| DropEdge2022.11 | 80.59 | |
| GCN2024.10 | 80.59 | |
| StructMAE2024.10 | 80.53 | |
| (s)GIN-epsilonneighborhood enlargement=true2019.05 | 80.51 | |
| WL-OA2024.06 | 80.5 | |
| AUG-MAE2025.12 | 80.5 | |
| WL-OAMethod Category=Kernels2025.10 | 80.5 | |
| Cluster-GT2024.10 | 80.43 | |
| HGLET + SVMInput Type=Combo, # Parameters=2562023.11 | 80.39 | |
| HGLET + SVMFeature Type=Combo2023.11 | 80.39 | |
| GraphMAEProtocol=Unsupervised representation learning2023.06 | 80.32 | |
| GraphMAE2024.10 | 80.32 | |
| GraphMAELearning Paradigm=Unsupervised2024.08 | 80.32 | |
| GraphMAE2025.12 | 80.3 | |
| GRAPHORMER2026.03 | 80.3 | |
| GIN-0neighborhood enlargement=false, epsilon=02019.05 | 80.2 | |
| GINinput=constant (X=1)2019.05 | 80.2 | |
| GIN-0Learning paradigm=Supervised Approaches2021.10 | 80.2 | |
| GIN-0Inference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 80.2 | |
| GIN2018.11 | 80.2 | |
| GINProtocol=Unsupervised representation learning2023.06 | 80.2 |