Graph Classification on MUTAG (test)
98.42AccuracyGraphTCL
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GraphTCL2025.12 | 98.42 | — | — | — | — | — | |
| PGOT12025.12 | 95.79 | — | — | — | — | — | |
| PXGL-GNNlearning_mode=supervised2025.12 | 94.87 | — | — | — | — | — | |
| OT-GNN2025.12 | 94.74 | — | — | — | — | — | |
| LiSABackbone=GIN2023.03 | 93.7 | — | — | — | — | — | |
| TFGW2025.12 | 93.68 | — | — | — | — | — | |
| SAGNNlearning_mode=supervised2025.12 | 93.24 | — | — | — | — | — | |
| GIN2025.12 | 93.13 | — | — | — | — | — | |
| CW Networkscategory=k-WL-based2025.12 | 92.7 | — | — | — | — | — | |
| SANlearning_mode=supervised2025.12 | 92.65 | — | — | — | — | — | |
| PATCHYSAN2018.10 | 92.6 | — | — | — | — | — | |
| PACHYSAN2019.11 | 92.6 | — | — | — | — | — | |
| PATCHY-SANcategory=GNNs2025.12 | 92.6 | — | — | — | — | — | |
| LGANresidual=false2025.12 | 92.5 | — | — | — | — | — | |
| FGSD2019.04 | 92.1 | — | — | — | — | — | |
| LGAN-resresidual=true2025.12 | 92 | — | — | — | — | — | |
| AttentionBackbone=GIN2023.03 | 91.7 | — | — | — | — | — | |
| GIN+GraphNormBackbone=GIN, Normalization=GraphNorm2020.09 | 91.6 | — | — | — | — | — | |
| TopoU-Netprotocol=Test@BestVal, pooling=global mean, rank path=0→1→2→1→02026.05 | 91.5 | — | — | — | — | — | |
| GINprotocol=Test@BestVal, pooling=global mean, rank path=0→1→2→1→02026.05 | 91.4 | — | — | — | — | — | |
| ICLlearning_mode=supervised2025.12 | 91.34 | — | — | — | — | — | |
| TopKPoolBackbone=GIN2023.03 | 91.3 | — | — | — | — | — | |
| GPRGNN2025.12 | 91.05 | — | — | — | — | — | |
| IRMBackbone=GIN2023.03 | 91 | — | — | — | — | — | |
| SpecMatch-CLtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 90.87 | — | — | — | — | — | |
| GFN2019.05 | 90.84 | — | — | — | — | — | |
| PPGNcategory=k-WL-based2025.12 | 90.6 | — | — | — | — | — | |
| CuCotraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 90.55 | — | — | — | — | — | |
| GIN+INSTANCENORMBackbone=GIN, Normalization=InstanceNorm2020.09 | 90.5 | — | — | — | — | — | |
| F-GINSupervision=Fully-supervised, Downstream Classifier=Linear2021.06 | 90.41 | — | — | — | — | — | |
| WL SUBTREEBackbone=WL SUBTREE2020.09 | 90.4 | — | — | — | — | — | |
| WL subtree2018.10 | 90.4 | — | — | — | — | — | |
| RetGK2019.04 | 90.3 | — | — | — | — | — | |
| RetGKgamma=12019.05 | 90.3 | — | — | — | — | — | |
| ERMBackbone=GIN2023.03 | 90.3 | — | — | — | — | — | |
| RetGKgamma=112019.05 | 90.1 | — | — | — | — | — | |
| TopoGCL2025.12 | 90.09 | — | — | — | — | — | |
| SUM-1-LAYERAggregation=SUM, Architecture=1-layer2018.10 | 90 | — | — | — | — | — | |
| GIN2019.04 | 90 | — | — | — | — | — | |
| GNTK2019.05 | 90 | — | — | — | — | — | |
| V-RexBackbone=GIN2023.03 | 90 | — | — | — | — | — | |
| GFN-light2019.05 | 89.89 | — | — | — | — | — | |
| AD-GCL-OPTSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 89.7 | — | — | — | — | — | |
| MSSGCLtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 89.68 | — | — | — | — | — | |
| DIRBackbone=GIN2023.03 | 89.5 | — | — | — | — | — | |
| GIN+BATCHNORMBackbone=GIN, Normalization=BatchNorm2020.09 | 89.4 | — | — | — | — | — | |
| SUM-MLP (GIN-0)Aggregation=SUM, Architecture=MLP2018.10 | 89.4 | — | — | — | — | — | |
| GIN2019.11 | 89.4 | — | — | — | — | — | |
| GIN2019.05 | 89.4 | — | — | — | — | — | |
| GINcategory=GNNs2025.12 | 89.4 | — | — | — | — | — | |
| S2GAElearning_mode=supervised2025.12 | 89.27 | — | — | — | — | — | |
| AD-GCL-FIXSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 89.25 | — | — | — | — | — | |
| NAD-GCL-OPTSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 89.05 | — | — | — | — | — | |
| InfoGraphtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 89.01 | — | — | — | — | — | |
| SimGRACEtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 89.01 | — | — | — | — | — | |
| SUM-MLP (GIN-ε)Aggregation=SUM, Architecture=MLP2018.10 | 89 | — | — | — | — | — | |
| PSCN2019.04 | 89 | — | — | — | — | — | |
| COMPGCN2019.11 | 89 | — | — | — | — | — | |
| PSCN2019.05 | 88.95 | — | — | — | — | — | |
| GIBBackbone=GIN2023.03 | 88.7 | — | — | — | — | — | |
| NAD-GCL-FIXSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 88.58 | — | — | — | — | — | |
| WKPI-kCvariant=kC2019.04 | 88.3 | — | — | — | — | — | |
| GraphCLSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 88.29 | — | — | — | — | — | |
| AWLBackbone=AWL2020.09 | 87.9 | — | — | — | — | — | |
| AWL2018.10 | 87.9 | — | — | — | — | — | |
| GCN2025.12 | 87.89 | — | — | — | — | — | |
| AWE2019.05 | 87.87 | — | — | — | — | — | |
| InfoGraphSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 87.71 | — | — | — | — | — | |
| JOAO V2training_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 87.67 | — | — | — | — | — | |
| RU-GINSupervision=Unsupervised, Downstream Classifier=Linear2021.06 | 87.61 | — | — | — | — | — | |
| SubGNNlearning_mode=supervised2025.12 | 87.52 | — | — | — | — | — | |
| DIFFPOOL2025.12 | 87.5 | — | — | — | — | — | |
| DGK2019.05 | 87.44 | — | — | — | — | — | |
| DGKtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 87.44 | — | — | — | — | — | |
| DGK2019.04 | 87.4 | — | — | — | — | — | |
| GAT2025.12 | 87.37 | — | — | — | — | — | |
| APPNP2025.12 | 87.37 | — | — | — | — | — | |
| JOAOtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 87.35 | — | — | — | — | — | |
| GCN2019.05 | 87.2 | — | — | — | — | — | |
| GIN-GUMPClass=Ours2024.03 | 87.11 | — | — | — | — | — | |
| att-Pooling-NLSFsArchitecture=Pooling2024.06 | 86.89 | — | — | — | — | — | |
| GraphCLtraining_regime=unsupervised, evaluation_protocol=10-fold cross-validation with linear SVM2025.11 | 86.8 | — | — | — | — | — | |
| DiffPoollearning_mode=supervised2025.12 | 86.72 | — | — | — | — | — | |
| CapsGNN2019.05 | 86.67 | — | — | — | — | — | |
| GRAPHSAGElearning_mode=supervised2025.12 | 86.35 | — | — | — | — | — | |
| 1-2-3 GNNcategory=k-WL-based2025.12 | 86.1 | — | — | — | — | — | |
| DGCNN2019.05 | 85.83 | — | — | — | — | — | |
| DGCNNBackbone=DGCNN2020.09 | 85.8 | — | — | — | — | — | |
| DGCNN2018.10 | 85.8 | — | — | — | — | — | |
| WKPI-KMvariant=KM2019.04 | 85.8 | — | — | — | — | — | |
| DGCNN2019.11 | 85.8 | — | — | — | — | — | |
| DGCNN2019.11 | 85.8 | — | — | — | — | — | |
| DGCNNcategory=GNNs2025.12 | 85.8 | — | — | — | — | — | |
| DropGIN2025.12 | 85.79 | — | — | — | — | — | |
| ChebyNet2025.12 | 85.79 | — | — | — | — | — | |
| MEAN-1-LAYER (GCN)Aggregation=MEAN, Architecture=1-layer2018.10 | 85.6 | — | — | — | — | — | |
| WL-Kernel2025.12 | 85.26 | — | — | — | — | — | |
| P-WL-UC2019.04 | 85.2 | — | — | — | — | — | |
| CCMambaTopological Abstraction=Rank-cell2026.01 | 85.11 | — | — | — | — | — | |
| MAX-1-LAYER (GraphSAGE)Aggregation=MAX, Architecture=1-layer2018.10 | 85.1 | — | — | — | — | — |