Node Classification on Computers (Accuracy %)
93.68Accuracy (%)Polynormer
Evaluation Results
| Method | Links | |
|---|---|---|
| Polynormer2026.07 | 93.68 | |
| DMbaGCNModel Type=Graph Mamba2025.11 | 92.49 | |
| X-LogSMaskConfiguration=ours2026.07 | 92.01 | |
| SGFormer2026.07 | 91.99 | |
| SSGCModel Type=Deep GNN2025.11 | 91.98 | |
| GPRGNNModel Type=Deep GNN2025.11 | 91.8 | |
| DMbaGCN w/o GCAMbaModel Type=Graph Mamba2025.11 | 91.74 | |
| SGCModel Type=GNN2025.11 | 91.62 | |
| Exphormer2026.07 | 91.47 | |
| NAGphormer2026.07 | 91.22 | |
| X-LogSMaskLayers=12026.07 | 91.21 | |
| GraphSAGE2026.07 | 91.2 | |
| GraphGPS2026.07 | 91.19 | |
| GTModel Type=Graph Transformer2025.11 | 91.18 | |
| SpexphormerModel Type=Graph Transformer2025.11 | 91.09 | |
| GCNModel Type=GNN2025.11 | 91.07 | |
| GOAT2026.07 | 90.96 | |
| GATModel Type=GNN2025.11 | 90.94 | |
| GAT2026.07 | 90.78 | |
| MbaGCNModel Type=Graph Mamba2025.11 | 90.39 | |
| SANModel Type=Graph Transformer2025.11 | 89.93 | |
| GCN2026.07 | 89.65 | |
| APPNPModel Type=Deep GNN2025.11 | 89.51 | |
| NodeFormer2026.07 | 86.98 | |
| GCNIIModel Type=Deep GNN2025.11 | 84.71 | |
| DMbaGCN w/o LSEMbaModel Type=Graph Mamba2025.11 | 84.18 | |
| S2GAEPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 65.59 | |
| LR-GMPPre-training Strategy=Edgepred-Gprompt2026.04 | 60.99 | |
| LR-GMPPre-training Strategy=Edgepred-GPPT2026.04 | 60.35 | |
| LR-GMPPre-training Strategy=SGCL2026.04 | 59.04 | |
| UniPromptPre-training Strategy=Edgepred-GPPT2026.04 | 58.24 | |
| LR-GMPPre-training Strategy=SUGRL2026.04 | 57.65 | |
| SCGFMPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 57.35 | |
| UniPromptPre-training Strategy=Edgepred-Gprompt2026.04 | 57.28 | |
| UniPromptPre-training Strategy=SGCL2026.04 | 56.39 | |
| GraphTOPPre-training Strategy=Edgepred-GPPT2026.04 | 55.65 | |
| UniPromptPre-training Strategy=SUGRL2026.04 | 53.97 | |
| GCOPEPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 53.7 | |
| GraphTOPPre-training Strategy=SGCL2026.04 | 53.14 | |
| GraphTOPPre-training Strategy=Edgepred-Gprompt2026.04 | 52.3 | |
| GraphTOPPre-training Strategy=SUGRL2026.04 | 52.26 | |
| GraphACLPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 51.48 | |
| FTPre-training Strategy=Edgepred-GPPT2026.04 | 50.98 | |
| EdgePromptPre-training Strategy=Edgepred-GPPT2026.04 | 50.3 | |
| EdgePromptPre-training Strategy=Edgepred-Gprompt2026.04 | 50.28 | |
| FTPre-training Strategy=SUGRL2026.04 | 50.18 | |
| FTPre-training Strategy=Edgepred-Gprompt2026.04 | 49.25 | |
| EdgePrompt+Pre-training Strategy=Edgepred-Gprompt2026.04 | 47.91 | |
| EdgePrompt+Pre-training Strategy=Edgepred-GPPT2026.04 | 47.91 | |
| EdgePromptPre-training Strategy=SGCL2026.04 | 46.99 | |
| EdgePrompt+Pre-training Strategy=SGCL2026.04 | 46.48 | |
| EdgePrompt+Pre-training Strategy=SUGRL2026.04 | 46.32 | |
| EdgePromptPre-training Strategy=SUGRL2026.04 | 44.93 | |
| GPFPre-training Strategy=Edgepred-GPPT2026.04 | 43.14 | |
| DGIPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 42.18 | |
| GPF-plusPre-training Strategy=SGCL2026.04 | 41.91 | |
| GPF-plusPre-training Strategy=Edgepred-Gprompt2026.04 | 41.33 | |
| GPF-plusPre-training Strategy=SUGRL2026.04 | 39.88 | |
| FTPre-training Strategy=SGCL2026.04 | 39.58 | |
| GPFPre-training Strategy=SGCL2026.04 | 39.35 | |
| GPFPre-training Strategy=SUGRL2026.04 | 38.44 | |
| GPF-plusPre-training Strategy=Edgepred-GPPT2026.04 | 38.18 | |
| All in OnePre-training Strategy=Edgepred-Gprompt2026.04 | 36.93 | |
| All in OnePre-training Strategy=SGCL2026.04 | 36.91 | |
| GPFPre-training Strategy=Edgepred-Gprompt2026.04 | 36.11 | |
| All in OnePre-training Strategy=Edgepred-GPPT2026.04 | 35.9 | |
| GITPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 35.76 | |
| All in OnePre-training Strategy=SUGRL2026.04 | 34.55 | |
| GINPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 32.52 | |
| GraphMAEPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 31.86 | |
| RiemannGFMPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 31.44 | |
| GATPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 31.01 | |
| GraphCLPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 30.44 | |
| GCNPre-training Domain=Trained on N1 (OOD), Evaluation protocol=5-shot2026.05 | 28 |