Node Classification on Minesweeper
99.03AccuracyGraphTARIF
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GraphTARIF2026.05 | 99.03 | — | |
| BuNN2026.05 | 98.99 | — | |
| M3Dphormer2026.05 | 98.27 | — | |
| GAT2026.05 | 97.86 | — | |
| GATRefinement=TS2026.05 | 97.86 | — | |
| GCNRefinement=TS2026.05 | 97.8 | — | |
| Polynormer2026.05 | 97.49 | — | |
| NT2026.04 | 97.42 | — | |
| SAGERefinement=TS2026.05 | 97.33 | — | |
| GCN2026.05 | 97.26 | — | |
| SAGE2026.05 | 97.09 | — | |
| GCN+ReP2026.05 | 96.05 | — | |
| CDE2026.04 | 95.5 | — | |
| GAT-sep2026.04 | 93.91 | — | |
| Dir-Poly2026.05 | 93.74 | — | |
| GraphSAGE2026.04 | 93.51 | — | |
| BloomGML2026.04 | 93.3 | — | |
| GAT2026.04 | 92.01 | — | |
| tGNN2026.04 | 91.93 | — | |
| SGFormer2026.05 | 91.42 | — | |
| GBK-GNN2026.04 | 90.85 | — | |
| GraphGPS2026.05 | 90.75 | — | |
| FSGNN2026.05 | 90.08 | — | |
| Specformer2026.05 | 89.93 | — | |
| GCNTraining Dimension=512, Supervised=true2023.10 | 89.75 | — | |
| GCN2026.04 | 89.75 | — | |
| H2GCNTraining Dimension=512, Supervised=true2023.10 | 89.71 | — | |
| H2GCN2026.04 | 89.71 | — | |
| H2GCN2026.05 | 89.71 | — | |
| JacobiConv2026.04 | 89.66 | — | |
| DNSD (diag)Map=diag, Adj=true, Odd=true, Gate=true, Best layer depth=L82026.05 | 89.4 | — | |
| DNSD (diag)Map=diag, Adj=true, Odd=true, Gate=false, Best layer depth=L82026.05 | 88.9 | — | |
| GCNEvaluation Setting=transductive2024.07 | 88.72 | — | |
| IPR-MPNN2024.05 | 88.7 | — | |
| DNSD (diag)Map=diag, Adj=false, Odd=true, Gate=true, Best layer depth=L82026.05 | 88.2 | — | |
| FAGCN2026.04 | 88.17 | — | |
| DNSD (diag)Map=diag, Adj=true, Odd=false, Gate=true, Best layer depth=L82026.05 | 88.1 | — | |
| NodeFormer2026.05 | 87.71 | — | |
| DNSD (diag)Map=diag, Adj=false, Odd=false, Gate=false, Best layer depth=L82026.05 | 87.6 | — | |
| GGCN2026.04 | 87.54 | — | |
| NSDMap=diag, Best layer depth=L82026.05 | 87.5 | — | |
| GHCEvaluation Setting=transductive2024.07 | 87.49 | — | |
| HMH (ours)2026.05 | 87.4 | — | |
| MPNNBest layer depth=L42026.05 | 87.4 | — | |
| Dir-GNN2026.04 | 87.05 | — | |
| TFE–GNN(sum)2026.05 | 86.85 | — | |
| UniFilter†2026.05 | 86.8 | — | |
| GCNEvaluation Setting=inductive2024.07 | 86.4 | — | |
| GPRGNNTraining Dimension=512, Supervised=true2023.10 | 86.24 | — | |
| GPRGNN2026.04 | 86.24 | — | |
| GPRGNN2026.05 | 86.24 | — | |
| GPS2024.02 | 86.2 | — | |
| NSDMap=full, Best layer depth=L42026.05 | 86.1 | — | |
| MVGRLRepresentation Dimension=512, RFF Projection=false2023.10 | 85.6 | — | |
| FIGURE_512Representation Dimension=512, RFF Projection=false2023.10 | 85.58 | — | |
| FiGUReTraining Dimension=512, RFF Projection=false2023.10 | 85.58 | — | |
| FIGURE_RFFRepresentation Dimension=32, RFF Projection=true2023.10 | 85.28 | — | |
| FiGURe-RFFTraining Dimension=32, RFF Projection=true2023.10 | 85.28 | — | |
| GHCEvaluation Setting=inductive2024.07 | 85.23 | — | |
| ESA2024.02 | 85.2 | — | |
| FIGURE_RRepresentation Dimension=128, RFF Projection=true2023.10 | 85.16 | — | |
| FiGURe-RFFTraining Dimension=128, RFF Projection=true2023.10 | 85.16 | — | |
| OptBasisGNN2026.05 | 84.8 | — | |
| SLOG(N)2026.05 | 84.4 | — | |
| GATEvaluation protocol=End-to-end2026.03 | 84.15 | — | |
| MeanGNNEvaluation protocol=End-to-end2026.03 | 84.06 | — | |
| GHMEvaluation Setting=transductive2024.07 | 84.02 | — | |
| GHMEvaluation Setting=inductive2024.07 | 83.88 | — | |
| ALL-IN2026.05 | 82.93 | — | |
| DGIRepresentation Dimension=512, RFF Projection=false2023.10 | 82.51 | — | |
| SUGRLRepresentation Dimension=512, RFF Projection=false2023.10 | 82.4 | — | |
| ChebNetII2026.05 | 82.3 | — | |
| M2M-GNN†2026.05 | 82.25 | — | |
| MTGCN2026.05 | 82.2 | — | |
| GCNModel=GCN, Training Source=Target dataset2026.04 | 81.12 | — | |
| GCN2026.05 | 81.12 | — | |
| GOBLINEvaluation protocol=Zero-shot2026.03 | 81 | — | |
| TS-MeanEvaluation protocol=Zero-shot2026.03 | 80.68 | — | |
| NodePFNModel=NodePFN, Training Source=Single pre-trained model2026.04 | 80.66 | — | |
| OrderedGNN2026.04 | 80.58 | — | |
| GraphAnyEvaluation protocol=Zero-shot2026.03 | 80.46 | — | |
| GraphAny (Cora)Model=GraphAny, Training Source=Cora2026.04 | 80.46 | — | |
| GraphAny2026.05 | 80.46 | — | |
| GraphAny (Arxiv)Model=GraphAny, Training Source=Arxiv2026.04 | 80.3 | — | |
| GATBest layer depth=L22026.05 | 80.3 | — | |
| GraphAny (Products)Model=GraphAny, Training Source=Products2026.04 | 80.27 | — | |
| GRACERepresentation Dimension=512, RFF Projection=false2023.10 | 80.22 | — | |
| GraphAny (Wisconsin)Model=GraphAny, Training Source=Wisconsin2026.04 | 80.13 | — | |
| PNA2024.02 | 80.1 | — | |
| GATModel=GAT, Training Source=Target dataset2026.04 | 80.08 | — | |
| MLPEvaluation Setting=inductive2024.07 | 80 | — | |
| MLPModel=MLP, Training Source=Target dataset2026.04 | 80 | — | |
| MLPBest layer depth=L22026.05 | 80 | — | |
| GINEvariant=base2024.05 | 79.9 | — | |
| GPRGNN2026.05 | 79.1 | — | |
| GCN2026.05 | 78.8 | — | |
| BernNet2026.05 | 78.8 | — | |
| GCNII2026.05 | 78.8 | — | |
| JacobiConv2026.05 | 78.8 | — | |
| GAT2026.05 | 78.7 | — |