Node Classification on Wiki-CS (test)
87.93AccuracyGraphSSR
Evaluation results
| Method | Links | |
|---|---|---|
| GraphSSRNumber of candidate categories=52026.03 | 87.93 | |
| Graph-R1Number of candidate categories=52026.03 | 86.89 | |
| D2MoEBackbone=GAT, Method Category=Ours2026.04 | 85.45 | |
| D2MoEBackbone=SAGE, Method Category=Ours2026.04 | 85.23 | |
| MOWST-2024Method Category=GRAPH MOE2026.04 | 85.13 | |
| ACMGCN-2022Method Category=HETERO2026.04 | 85.1 | |
| FSGNN-2022Method Category=HETERO2026.04 | 85.1 | |
| NODEMOE-2024Method Category=GRAPH MOE2026.04 | 85.08 | |
| GPRGNN-2021Method Category=HETERO2026.04 | 84.84 | |
| D2MoEBackbone=GCN, Method Category=Ours2026.04 | 84.81 | |
| SAGEMethod Category=VANILLA2026.04 | 84.75 | |
| H2GCN-2020Method Category=HETERO2026.04 | 84.62 | |
| EXPHORMER-2023Method Category=GT2026.04 | 84.57 | |
| FAGCN-2021Method Category=HETERO2026.04 | 84.47 | |
| NAGFORMER-2023Method Category=GT2026.04 | 84.31 | |
| DAMOE-2025Method Category=GRAPH MOE2026.04 | 84.04 | |
| GATMethod Category=VANILLA2026.04 | 83.99 | |
| GCNMethod Category=VANILLA2026.04 | 83.8 | |
| GMOE-2023Method Category=GRAPH MOE2026.04 | 83.65 | |
| DIFFORMER-2023Method Category=GT2026.04 | 83.54 | |
| ANS-GT-2022Method Category=GT2026.04 | 83.27 | |
| UltraTAG-SSparsity=0%2025.04 | 83.05 | |
| SGFORMER-2024Method Category=GT2026.04 | 82.67 | |
| MOSCAT-2025Method Category=GRAPH MOE2026.04 | 81.4 | |
| ENGINESparsity=0%2025.04 | 81.38 | |
| GAT2026.01 | 81.01 | |
| GOFA-TNumber of candidate categories=52026.03 | 80.93 | |
| GCNBBackbone=2-layer GCN2025.05 | 80.75 | |
| GraphPatcherBackbone=2-layer GCN2025.05 | 80.64 | |
| SAGE2026.01 | 80.57 | |
| TUNEUPBackbone=2-layer GCN2025.05 | 80.56 | |
| GOFA-FNumber of candidate categories=52026.03 | 80.52 | |
| DropMessageBackbone=2-layer GCN2025.05 | 80.3 | |
| GCNBackbone=2-layer GCN2025.05 | 80.26 | |
| FAFselection=best validation2026.01 | 80.25 | |
| DropEdgeBackbone=2-layer GCN2025.05 | 80.22 | |
| TAPESparsity=0%2025.04 | 80.17 | |
| STEM-GNN2026.02 | 80.11 | |
| DropNodeBackbone=2-layer GCN2025.05 | 80.11 | |
| GCN2026.01 | 80.06 | |
| MLPMethod Category=VANILLA2026.04 | 79.57 | |
| GraphSSRNumber of candidate categories=102026.03 | 79.4 | |
| VANILLA GTMethod Category=GT2026.04 | 79.05 | |
| CaNet2026.02 | 78.88 | |
| Graph-R1Number of candidate categories=102026.03 | 78.68 | |
| MARIO2026.02 | 78.6 | |
| GFT2026.02 | 78.4 | |
| GraphMAE2026.02 | 77.61 | |
| UltraTAG-SSparsity=50%2025.04 | 77.45 | |
| GAT2026.02 | 76.78 | |
| TFEGNN2026.02 | 76.74 | |
| UniGraphLMLearning Setting=multi-domain multi-task, Fine-tuning=None (joint instruction tuning only)2026.05 | 76.65 | |
| GIANT2026.02 | 76.56 | |
| BGRL2026.02 | 76.53 | |
| SimTeGSparsity=0%2025.04 | 76.32 | |
| MLPBackbone=MLP2025.05 | 75.98 | |
| DGI2026.02 | 75.32 | |
| GCN2026.02 | 75.28 | |
| GraphSAGESparsity=0%2025.04 | 75.16 | |
| CANEBackbone=GAT, Annotator=gpt-3.5-turbo2026.05 | 74.91 | |
| GLEMSparsity=0%2025.04 | 74.83 | |
| GraphMETRO2026.02 | 74.59 | |
| TAPESparsity=50%2025.04 | 73.62 | |
| GCNSparsity=0%2025.04 | 73.27 | |
| LOCLEBackbone=GCN, Annotator=gpt-3.5-turbo2026.05 | 72.73 | |
| CANEBackbone=GCN, Annotator=gpt-3.5-turbo2026.05 | 72.7 | |
| RoBERTaSparsity=0%2025.04 | 72.12 | |
| ENGINESparsity=50%2025.04 | 71.72 | |
| BERTSparsity=0%2025.04 | 71.7 | |
| GOFA-TNumber of candidate categories=102026.03 | 71.17 | |
| Linear2026.02 | 70.36 | |
| GCNIISparsity=0%2025.04 | 70.12 | |
| LLaGALearning Setting=multi-domain multi-task, Fine-tuning=None (joint instruction tuning only)2026.05 | 69.16 | |
| LOCLEBackbone=GAT, Annotator=gpt-3.5-turbo2026.05 | 68.86 | |
| GOFA-FNumber of candidate categories=102026.03 | 68.84 | |
| DMABackbone=GAT, Annotator=gpt-3.5-turbo2026.05 | 68.73 | |
| GATSparsity=0%2025.04 | 68.72 | |
| DeBERTaSparsity=0%2025.04 | 68.18 | |
| LLM-GNN (GP)Backbone=GCN, Annotator=gpt-3.5-turbo2026.05 | 67.58 | |
| LLM-GNN (FP)Backbone=GCN, Annotator=gpt-3.5-turbo2026.05 | 67.31 | |
| GLEMSparsity=50%2025.04 | 67.07 | |
| GOFALearning Setting=multi-domain multi-task, Fine-tuning=None (joint instruction tuning only)2026.05 | 66.77 | |
| DMABackbone=GCN, Annotator=gpt-3.5-turbo2026.05 | 66.42 | |
| UltraTAG-SSparsity=80%2025.04 | 65.6 | |
| SimTeGSparsity=50%2025.04 | 65.34 | |
| LLM-GNN (FP)Backbone=GAT, Annotator=gpt-3.5-turbo2026.05 | 65.33 | |
| LLM-GNN (RIM)Backbone=GCN, Annotator=gpt-3.5-turbo2026.05 | 64.79 | |
| GATSparsity=50%2025.04 | 64.6 | |
| GraphSAGESparsity=50%2025.04 | 64.56 | |
| GCNSparsity=50%2025.04 | 64.12 | |
| LLM-GNN (RIM)Backbone=GAT, Annotator=gpt-3.5-turbo2026.05 | 64.06 | |
| UniGraphLMSetting=Cross-domain, Zero-shot transfer=true, Trained on=Arxiv2026.05 | 63.93 | |
| LLM-GNN (GP)Backbone=GAT, Annotator=gpt-3.5-turbo2026.05 | 63.74 | |
| GCNIISparsity=50%2025.04 | 63.62 | |
| SA2GFMshots=5-shot, attack_type=non-targeted feature attack, lambda=0.42025.11 | 63.5 | |
| MLPSparsity=0%2025.04 | 62.46 | |
| Vicuna-7BLearning Setting=multi-domain multi-task, Fine-tuning=None (joint instruction tuning only)2026.05 | 62.34 | |
| MDGFMshots=5-shot, attack_type=non-targeted feature attack, lambda=0.42025.11 | 61.5 | |
| UniGraphNumber of candidate categories=52026.03 | 60.23 | |
| TAPESparsity=80%2025.04 | 59.83 |