Graph Classification on HIV
0.832ROC-AUC<SOGk>
Evaluation Results
| Method | Links | |
|---|---|---|
| <SOGk>Backbone=LLaMA2-7B2026.02 | 0.832 | |
| LEAPPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7942 | |
| SUPTsoftPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7931 | |
| LEAPPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.7929 | |
| FTPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7915 | |
| SUPThardPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7915 | |
| GPF-plusPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7912 | |
| GPFPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7904 | |
| RELIEFPre-training Strategy=ContextPred, Shot Scenario=full-shot2025.12 | 0.7897 | |
| SUPTsoftPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.7895 | |
| GPFPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.7891 | |
| GPF-plusPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.7877 | |
| SUPThardPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.786 | |
| S-CGIBDomain Adaptation=true2024.12 | 0.7833 | |
| FTPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.7831 | |
| RELIEFPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.7817 | |
| SUPTsoftPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.7815 | |
| SimSGT2024.12 | 0.7813 | |
| FTPre-training Strategy=AttrMasking, Shot Scenario=full-shot2025.12 | 0.7813 | |
| LEAPPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.7812 | |
| MoAMa2024.12 | 0.7811 | |
| LLaMA2-7BProtocol=LoRA SFT2026.02 | 0.781 | |
| SUPThardPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.7806 | |
| RELIEFPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.778 | |
| LEAPPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.7776 | |
| GPF-plusPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.777 | |
| GPF-plusPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.777 | |
| FTPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.7758 | |
| JOAO2024.12 | 0.7755 | |
| MGSSL2024.12 | 0.7745 | |
| SUPTsoftPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.7742 | |
| S-CGIBDomain Adaptation=false2024.12 | 0.7741 | |
| SUPThardPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.7731 | |
| GPFPre-training Strategy=GCL, Shot Scenario=full-shot2025.12 | 0.7722 | |
| JOAOv22024.12 | 0.7713 | |
| GPFPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.7705 | |
| JOAOPre-Train dataset=ZINC 2M2025.11 | 0.769 | |
| RELIEFPre-training Strategy=Infomax, Shot Scenario=full-shot2025.12 | 0.7684 | |
| MICRO-Graph2024.12 | 0.7673 | |
| GTTok.=Comb., Mode=Trunc.2026.05 | 0.7644 | |
| EdgePredPre-Train dataset=ZINC 2M2025.11 | 0.763 | |
| SpecMatch-CLPre-Train dataset=ZINC 2M2025.11 | 0.7625 | |
| GraphFP2024.12 | 0.7571 | |
| <SOGk>Backbone=LLaMA3-3B2026.02 | 0.757 | |
| ContextPredPre-Train dataset=ZINC 2M2025.11 | 0.756 | |
| No Pre-TrainPre-Train dataset=None2025.11 | 0.755 | |
| GTTok.=Lap, Mode=Pad.2026.05 | 0.7542 | |
| AttrMaskingPre-Train dataset=ZINC 2M2025.11 | 0.753 | |
| SimGRACEPre-Train dataset=ZINC 2M2025.11 | 0.752 | |
| GraphCLPre-Train dataset=ZINC 2M2025.11 | 0.751 | |
| GROVE2024.12 | 0.7504 | |
| DeepSet2026.05 | 0.7482 | |
| GFT + RGVQRegime=Pre-trained and Fine-tuned2025.08 | 0.741 | |
| GTTok.=RW, Mode=Trunc.2026.05 | 0.7408 | |
| GraphLoG2024.12 | 0.7376 | |
| Infomax2024.12 | 0.7355 | |
| GraphCL2024.12 | 0.7355 | |
| FedGALA2026.01 | 0.733 | |
| GIN2026.05 | 0.732 | |
| GTTok.=Lap, Mode=Trunc.2026.05 | 0.7315 | |
| GTTok.=Comb., Mode=Pad.2026.05 | 0.731 | |
| GFT + ResetRegime=Pre-trained and Fine-tuned2025.08 | 0.7279 | |
| GTTok.=RW, Mode=Pad.2026.05 | 0.7275 | |
| AttrMasking2024.12 | 0.7271 | |
| MolCA-GSBackbone=Non-LLM2026.02 | 0.727 | |
| GFT + EMARegime=Pre-trained and Fine-tuned2025.08 | 0.7239 | |
| GFTRegime=Pre-trained and Fine-tuned2025.08 | 0.7216 | |
| ContextPred2024.12 | 0.7204 | |
| GFT + APRegime=Pre-trained and Fine-tuned2025.08 | 0.7186 | |
| GTTok.=Adj, Mode=Pad.2026.05 | 0.7131 | |
| GPF-AttrMaskingBackbone=Non-LLM2026.02 | 0.713 | |
| GTTok.=Adj, Mode=Trunc.2026.05 | 0.713 | |
| GFT + PTRegime=Pre-trained and Fine-tuned2025.08 | 0.7101 | |
| EdgePred2024.12 | 0.7055 | |
| FedBook2026.01 | 0.701 | |
| FedGFM+2026.01 | 0.697 | |
| OFAfederated adoption of GFM=true2026.01 | 0.6954 | |
| UniGraphfederated adoption of GFM=true2026.01 | 0.684 | |
| GQTfederated adoption of GFM=true2026.01 | 0.6832 | |
| GFTfederated adoption of GFM=true2026.01 | 0.6757 | |
| GCFL+2026.01 | 0.6734 | |
| LEAPPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6715 | |
| RELIEFPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6704 | |
| GPF-plusPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6691 | |
| GINRegime=Supervised2025.08 | 0.6686 | |
| SUPThardPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6681 | |
| GraphCLIPfederated adoption of GFM=true2026.01 | 0.6675 | |
| SUPTsoftPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6659 | |
| GFT + SimVQRegime=Pre-trained and Fine-tuned2025.08 | 0.6657 | |
| GPFPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6651 | |
| R-GFMShots=5-shot2026.05 | 0.6648 | |
| FedStar2026.01 | 0.662 | |
| MOON2026.01 | 0.6596 | |
| GPT-4Protocol=Zero-shot2026.02 | 0.659 | |
| FTPre-training Strategy=Infomax, Shot Scenario=50-shot2025.12 | 0.6587 | |
| R-GFMShots=3-shot2026.05 | 0.6557 | |
| GATRegime=Supervised2025.08 | 0.6554 | |
| GIANTRegime=Self-supervised2025.08 | 0.6544 | |
| GCOPEShots=5-shot2026.05 | 0.6536 | |
| GCNRegime=Supervised2025.08 | 0.6484 |