Node Classification on Photo (test)
96.13Mean AccuracyDirGNN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DirGNN2024.10 | 96.13 | — | |
| Co-GNN2024.10 | 95.95 | — | |
| FLODE2024.10 | 95.93 | — | |
| CoEDedge direction learning=false2024.10 | 95.83 | — | |
| D2MoEBackbone=GAT, Method Category=Ours2026.04 | 95.79 | — | |
| SignGT2023.10 | 95.68 | — | |
| D2MoEBackbone=SAGE, Method Category=Ours2026.04 | 95.64 | — | |
| H2GCN-2020Method Category=HETERO2026.04 | 95.59 | — | |
| D2MoEBackbone=GCN, Method Category=Ours2026.04 | 95.59 | — | |
| NODEMOE-2024Method Category=GRAPH MOE2026.04 | 95.53 | — | |
| FSGNN-2022Method Category=HETERO2026.04 | 95.5 | — | |
| GPRGNN2023.10 | 95.49 | — | |
| MOWST-2024Method Category=GRAPH MOE2026.04 | 95.49 | — | |
| MOSCAT-2025Method Category=GRAPH MOE2026.04 | 95.48 | — | |
| Chung2024.10 | 95.47 | — | |
| GPRGNN-2021Method Category=HETERO2026.04 | 95.44 | — | |
| ACMGCN-2022Method Category=HETERO2026.04 | 95.42 | — | |
| EXPHORMER-2023Method Category=GT2026.04 | 95.42 | — | |
| SAGEMethod Category=VANILLA2026.04 | 95.41 | — | |
| DIFFORMER-2023Method Category=GT2026.04 | 95.41 | — | |
| HiGCNOrder=22023.09 | 95.33 | — | |
| GraphSAGECategory=GNNs2026.06 | 95.31 | — | |
| NodeFormer2023.10 | 95.27 | — | |
| HiGCNOrder=12023.09 | 95.22 | — | |
| GloGNN2023.10 | 95.17 | — | |
| SGC2023.10 | 95.11 | — | |
| PPRGoCategory=GNNs2026.06 | 95.09 | — | |
| GraphGPS2025.12 | 95.06 | — | |
| Sheaf2024.10 | 95.01 | — | |
| GAT2023.10 | 94.98 | — | |
| GophormerCategory=GTs2026.06 | 94.98 | — | |
| NAGFORMER-2023Method Category=GT2026.04 | 94.95 | — | |
| SANCategory=GTs2026.06 | 94.85 | — | |
| GRAPH CASCADESVariant=GPS-TAS2026.06 | 94.84 | — | |
| GPS-TASBackbone=GraphGPS, Cascade Strategy=TAS2026.06 | 94.84 | — | |
| GPS-TASMethod Category=GNN-based Architectures, Rewiring=TAS2026.06 | 94.84 | — | |
| SPGCLTraining Data=X, A2026.06 | 94.83 | — | |
| GRAND+Category=GNNs2026.06 | 94.82 | — | |
| GCN2023.10 | 94.76 | — | |
| Space OptimumMaximum refinement budget=1002025.07 | 94.75 | — | |
| M-DESIGNMaximum refinement budget=1002025.07 | 94.75 | — | |
| GPS-MASBackbone=GraphGPS, Cascade Strategy=MAS2026.06 | 94.71 | — | |
| GPS-MASMethod Category=GNN-based Architectures, Rewiring=MAS2026.06 | 94.71 | — | |
| GraphGPSCategory=GTs2026.06 | 94.64 | — | |
| GraphGPSBackbone=GraphGPS, Cascade Strategy=None2026.06 | 94.64 | — | |
| GraphGPSMethod Category=GNN-based Architectures2026.06 | 94.64 | — | |
| AutoTransferMaximum refinement budget=1002025.07 | 94.62 | — | |
| NAG-TASBackbone=NAG, Cascade Strategy=TAS2026.06 | 94.61 | — | |
| NAG-TASMethod Category=Tokenized Sparse-Attention GTs, Rewiring=TAS2026.06 | 94.61 | — | |
| DesiGNNMaximum refinement budget=1002025.07 | 94.6 | — | |
| SAGE2024.10 | 94.59 | — | |
| GraphSAGE2025.12 | 94.59 | — | |
| FAGCN2023.10 | 94.56 | — | |
| MagNet2024.10 | 94.54 | — | |
| KBGMaximum refinement budget=1002025.07 | 94.53 | — | |
| Specformer2023.10 | 94.51 | — | |
| ANS-GT-2022Method Category=GT2026.04 | 94.51 | — | |
| GMOE-2023Method Category=GRAPH MOE2026.04 | 94.51 | — | |
| GGCN2024.10 | 94.5 | — | |
| GATCategory=GNNs2026.06 | 94.49 | — | |
| Auto-GNNMaximum refinement budget=1002025.07 | 94.46 | — | |
| SGFORMER-2024Method Category=GT2026.04 | 94.46 | — | |
| EAMaximum refinement budget=1002025.07 | 94.45 | — | |
| FAGCN-2021Method Category=HETERO2026.04 | 94.44 | — | |
| GATMethod Category=VANILLA2026.04 | 94.42 | — | |
| HiGCNOrder=32023.09 | 94.4 | — | |
| DAMOE-2025Method Category=GRAPH MOE2026.04 | 94.39 | — | |
| GraphNASMaximum refinement budget=1002025.07 | 94.38 | — | |
| RLMaximum refinement budget=1002025.07 | 94.37 | — | |
| graph NEEmbedding dimension=1282025.03 | 94.3 | — | |
| graph NEτEmbedding dimension=1282025.03 | 94.3 | — | |
| RandomMaximum refinement budget=1002025.07 | 94.28 | — | |
| NAG-MASBackbone=NAG, Cascade Strategy=MAS2026.06 | 94.27 | — | |
| NAG-MASMethod Category=Tokenized Sparse-Attention GTs, Rewiring=MAS2026.06 | 94.27 | — | |
| GCNMethod Category=VANILLA2026.04 | 94.26 | — | |
| ExphormerCategory=GTs2026.06 | 94.25 | — | |
| VCR-TASBackbone=VCR, Cascade Strategy=TAS2026.06 | 94.22 | — | |
| VCR-TASMethod Category=Tokenized Sparse-Attention GTs, Rewiring=TAS2026.06 | 94.22 | — | |
| GCN-TASBackbone=GCN, Cascade Strategy=TAS2026.06 | 94.17 | — | |
| GCN-TASMethod Category=GNN-based Architectures, Rewiring=TAS2026.06 | 94.17 | — | |
| HiGCNOrder=42023.09 | 94.1 | — | |
| All featuresselection_ratio=100%, backbone=GNN2025.10 | 94.04 | — | |
| CR-TASBackbone=CR-Graphormer, Cascade Strategy=TAS2026.06 | 93.95 | — | |
| CR-TASMethod Category=Tokenized Sparse-Attention GTs, Rewiring=TAS2026.06 | 93.95 | — | |
| SGRLTraining Data=X, A2026.06 | 93.95 | — | |
| NAGCategory=GTs2026.06 | 93.94 | — | |
| NAGBackbone=NAG, Cascade Strategy=None2026.06 | 93.94 | — | |
| NAGMethod Category=Tokenized Sparse-Attention GTs2026.06 | 93.94 | — | |
| StrGCLTraining Data=X, A2026.06 | 93.94 | — | |
| GCN-MASBackbone=GCN, Cascade Strategy=MAS2026.06 | 93.91 | — | |
| GCN-MASMethod Category=GNN-based Architectures, Rewiring=MAS2026.06 | 93.91 | — | |
| GCNCategory=GNNs2026.06 | 93.9 | — | |
| GCNBackbone=GCN, Cascade Strategy=None2026.06 | 93.9 | — | |
| GCNMethod Category=GNN-based Architectures2026.06 | 93.9 | — | |
| VCR-MASBackbone=VCR, Cascade Strategy=MAS2026.06 | 93.88 | — | |
| VCR-MASMethod Category=Tokenized Sparse-Attention GTs, Rewiring=MAS2026.06 | 93.88 | — | |
| GAT2024.10 | 93.87 | — | |
| GAT2025.12 | 93.87 | — | |
| GPRGNN2023.09 | 93.85 | — | |
| ChebNet2023.09 | 93.77 | — |