Graph Classification on NCI1
95.17AccuracyUAT-T
Evaluation results
| Method | Links | ||
|---|---|---|---|
| UAT-Tperturbation_type=targeted adversarial2026.03 | 95.17 | — | |
| Ba-Logic2026.03 | 94.38 | — | |
| CDATevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 93.5 | — | |
| UAT-Uperturbation_type=uniform adversarial2026.03 | 93.46 | — | |
| δ-2-LWL+Type=Local Kernel2019.04 | 91.4 | — | |
| ETevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 90.1 | — | |
| ESA2024.02 | 87.8 | — | |
| WKPI-KM2019.04 | 87.5 | — | |
| WKPIConfig Category=Top NT, Landmarks (K)=200, Evaluation Protocol=5 seeds x 10 folds2026.05 | 87.5 | — | |
| WKPI (k-means)evaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 87.5 | — | |
| WL-OA2024.06 | 86.3 | — | |
| WL-OACategory=Kernels2025.10 | 86.3 | — | |
| IPR-MPNN2024.05 | 86.2 | — | |
| WL-OAKernel Type=Weisfeiler-Lehman Optimal Assignment2016.06 | 86.1 | — | |
| WL-OAedge types=annotated only2018.11 | 86.1 | — | |
| WL-OAMethod category=Kernel2018.10 | 86.1 | — | |
| WL-OA2018.11 | 86.1 | — | |
| ShareGNNEvaluation Protocol=Standard eval.2026.05 | 86.1 | — | |
| WL kernelType=Graph kernel2023.06 | 86 | — | |
| WL KERNEL2024.05 | 86 | — | |
| WL kernelCategory=Graph Kernel Methods, Validation=Tenfold cross-validation2024.12 | 86 | — | |
| WL kernel2026.05 | 86 | — | |
| WLKernel Type=Weisfeiler-Lehman Convolution2016.06 | 85.8 | — | |
| PXGL-GNN2025.12 | 85.78 | — | |
| P-WL-UC2019.04 | 85.6 | — | |
| WL-PM2024.06 | 85.6 | — | |
| PR-MPNN2024.05 | 85.6 | — | |
| WL-PMCategory=Kernels2025.10 | 85.6 | — | |
| NCWalpha=1000, beta=02022.05 | 85.5 | — | |
| NCWWL2022.05 | 85.5 | — | |
| PDF2023.05 | 85.47 | — | |
| WL2019.04 | 85.4 | — | |
| WL2018.11 | 85.4 | — | |
| WL subtree2026.05 | 85.4 | — | |
| WLEvaluation Protocol=Standard eval.2026.05 | 85.4 | — | |
| WL2022.05 | 85.3 | — | |
| CIN++Type=Topological neural network2023.06 | 85.3 | — | |
| CIN++2024.05 | 85.3 | — | |
| WL-VH2024.06 | 85.2 | — | |
| CORE-WL-VH2024.06 | 85.2 | — | |
| WL-VHCategory=Kernels2025.10 | 85.2 | — | |
| CORE-WL-VHCategory=Kernels2025.10 | 85.2 | — | |
| GAT2024.02 | 85.1 | — | |
| PINEvaluation Protocol=Standard eval.2026.05 | 85.1 | — | |
| PNA2024.02 | 85 | — | |
| GPS2024.02 | 85 | — | |
| WLOAType=Baseline2019.04 | 84.9 | — | |
| Norm-GNevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 84.9 | — | |
| GIN2024.02 | 84.8 | — | |
| δ-2-LWLType=Local Kernel2019.04 | 84.7 | — | |
| NH2024.06 | 84.7 | — | |
| CWN2026.05 | 84.7 | — | |
| NHCategory=Kernels2025.10 | 84.7 | — | |
| WLedge types=annotated only2018.11 | 84.6 | — | |
| RetGK2019.04 | 84.5 | — | |
| WKPI-KC2019.04 | 84.5 | — | |
| RetGk2018.11 | 84.5 | — | |
| CANType=Topological neural network2023.06 | 84.5 | — | |
| CAN2024.05 | 84.5 | — | |
| DropGIN2024.02 | 84.3 | — | |
| GraphGPS2026.07 | 84.21 | — | |
| 1-WLType=Baseline2019.04 | 84.2 | — | |
| GCN2024.02 | 84.2 | — | |
| GNTKCategory=Graph Kernel Methods, Validation=Tenfold cross-validation2024.12 | 84.2 | — | |
| AIA2022.11 | 84.12 | — | |
| GIC2018.11 | 84.08 | — | |
| GIN + GRANOLAevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 84 | — | |
| VEPMLearning Paradigm=Supervised/Semi-supervised2024.08 | 83.9 | — | |
| CAL2022.11 | 83.89 | — | |
| M-Mixup2022.11 | 83.89 | — | |
| (s)GIN-0neighborhood enlargement=true, epsilon=02019.05 | 83.85 | — | |
| ECCedge types=annotated only2018.11 | 83.8 | — | |
| sCWNrPeak GPU Memory (MB)=262026.05 | 83.8 | — | |
| OPS-GFSadaptation=adapted from image domain2026.03 | 83.71 | — | |
| structure2vecedge types=annotated only2018.11 | 83.7 | — | |
| CINType=Topological neural network2023.06 | 83.6 | — | |
| GIN + MedianBackbone=GIN, Activation=Median2024.07 | 83.6 | — | |
| CIN2024.05 | 83.6 | — | |
| CINValidation=Tenfold cross-validation2024.12 | 83.6 | — | |
| SNNValidation=Tenfold cross-validation, Variant=SNN(α, (1, 2)), γ=0.52024.12 | 83.6 | — | |
| CINEvaluation Protocol=Standard eval.2026.05 | 83.6 | — | |
| CINPeak GPU Memory (MB)=422026.05 | 83.6 | — | |
| GSNType=Graph neural network2023.06 | 83.5 | — | |
| GIN + LeakyReLUBackbone=GIN, Activation=LeakyReLU2024.07 | 83.5 | — | |
| GIN + GeLUBackbone=GIN, Activation=GeLU2024.07 | 83.5 | — | |
| GSN2024.05 | 83.5 | — | |
| GSNValidation=Tenfold cross-validation2024.12 | 83.5 | — | |
| GSNEvaluation Protocol=Standard eval.2026.05 | 83.5 | — | |
| ICL2025.12 | 83.45 | — | |
| Multigraph ChebNetedge types=annotated and learned, pooling=global max pooling2018.11 | 83.4 | — | |
| δ-3-LWLType=Local Kernel2019.04 | 83.4 | — | |
| GIN + SwishBackbone=GIN, Activation=Swish2024.07 | 83.4 | — | |
| GIN + DIGRAFBackbone=GIN, Activation=DIGRAF, Adaptive=true2024.07 | 83.4 | — | |
| GIN + ELUBackbone=GIN, Activation=ELU2024.07 | 83.3 | — | |
| GIN + MaxoutBackbone=GIN, Activation=Maxout2024.07 | 83.3 | — | |
| GIN + MaxBackbone=GIN, Activation=Max2024.07 | 83.3 | — | |
| PPGNSType=Graph neural network2023.06 | 83.2 | — | |
| GIN + TanhBackbone=GIN, Activation=Tanh2024.07 | 83.2 | — | |
| PPGNS2024.05 | 83.2 | — | |
| GraphMAE + COREmasking process=GraphMAE2025.12 | 83.2 | — |