Molecular Property Prediction on MUV (test)
84.67ROC-AUCSUPTsoft
Evaluation Results
| Method | Links | |
|---|---|---|
| SUPTsoftPre-training Strategy=ContextPred2024.02 | 84.67 | |
| SUPThardPre-training Strategy=ContextPred2024.02 | 84.54 | |
| GPF-plusPre-training Strategy=ContextPred2022.09 | 84.48 | |
| GPF-plusPre-training Strategy=ContextPred2024.02 | 84.4 | |
| GPFPre-training Strategy=ContextPred2022.09 | 84.34 | |
| GPFPre-training Strategy=ContextPred2024.02 | 84.22 | |
| GPF-plusPre-training Strategy=EdgePred2022.09 | 83.13 | |
| SUPTsoftPre-training Strategy=EdgePred2024.02 | 82.95 | |
| SUPThardPre-training Strategy=EdgePred2024.02 | 82.94 | |
| GPFPre-training Strategy=EdgePred2022.09 | 82.86 | |
| GPF-plusPre-training Strategy=EdgePred2024.02 | 82.74 | |
| GPFPre-training Strategy=EdgePred2024.02 | 82.61 | |
| FTPre-training Strategy=ContextPred2022.09 | 82.36 | |
| FTPre-training Strategy=ContextPred2024.02 | 82.36 | |
| SPD-Sheaf2026.04 | 82.3 | |
| SMPT2026.04 | 82.2 | |
| GPFPre-training Strategy=AttrMasking2022.09 | 82.17 | |
| Uni-Mol2026.04 | 82.1 | |
| GPPT (w/o ol)Pre-training Strategy=EdgePred2022.09 | 82.06 | |
| GPPTw/o olPre-training Strategy=EdgePred2024.02 | 82.06 | |
| SUPTsoftPre-training Strategy=AttrMasking2024.02 | 81.93 | |
| SUPThardPre-training Strategy=AttrMasking2024.02 | 81.81 | |
| Hierarchical Inter-Message Passinglayers=2 or 32020.06 | 81.8 | |
| FTPre-training Strategy=AttrMasking2022.09 | 81.78 | |
| FTPre-training Strategy=AttrMasking2024.02 | 81.78 | |
| GEM2026.04 | 81.7 | |
| SUPTsoftPre-training Strategy=Infomax2024.02 | 81.59 | |
| SUPThardPre-training Strategy=Infomax2024.02 | 81.51 | |
| GPFPre-training Strategy=AttrMasking2024.02 | 81.44 | |
| FTPre-training Strategy=Infomax2022.09 | 81.42 | |
| FTPre-training Strategy=Infomax2024.02 | 81.42 | |
| GPF-plusPre-training Strategy=Infomax2022.09 | 81.33 | |
| PretrainGNN2026.04 | 81.3 | |
| GPF-plusPre-training Strategy=AttrMasking2024.02 | 81.27 | |
| GPF-plusPre-training Strategy=AttrMasking2022.09 | 81.12 | |
| MoleBERT+GFSEBackbone=MoleBERT, SE=GFSE2025.04 | 80.5 | |
| GPFPre-training Strategy=Infomax2022.09 | 80.43 | |
| MoleBERT+RWSEBackbone=MoleBERT, SE=RWSE2025.04 | 80.4 | |
| GPFPre-training Strategy=Infomax2024.02 | 80.39 | |
| GPF-plusPre-training Strategy=Infomax2024.02 | 79.94 | |
| NGF2020.06 | 79.8 | |
| SSP2025.04 | 79.8 | |
| FTPre-training Strategy=EdgePred2022.09 | 79.67 | |
| FTPre-training Strategy=EdgePred2024.02 | 79.67 | |
| MolCLR2026.04 | 79.6 | |
| GIN-E2020.06 | 79.57 | |
| MGSSL (DFS)Backbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 79.5 | |
| RP-NGF2020.06 | 79.4 | |
| MGSSL (BFS)Backbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 78.8 | |
| MoleBERTBackbone=MoleBERT, Training=Pre-trained2025.04 | 78.6 | |
| D-MPNN2026.04 | 78.6 | |
| AttentiveFP2026.04 | 78.6 | |
| GraphMAE+GFSEBackbone=GraphMAE, SE=GFSE2025.04 | 78.1 | |
| Attribute maskingBackbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 77.9 | |
| GroverBackbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 77.8 | |
| GINE+GFSEBackbone=GINE, SE=GFSE2025.04 | 77.7 | |
| GraphMAE+RWSEBackbone=GraphMAE, SE=RWSE2025.04 | 77.7 | |
| GraphMVP2026.04 | 77.7 | |
| GPT-GNNBackbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 77 | |
| GINE+LapPEBackbone=GINE, SE=LapPE2025.04 | 77 | |
| N-GramClassifier=RF2026.04 | 76.9 | |
| InfomaxBackbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 76.5 | |
| GINE+RWSEBackbone=GINE, SE=RWSE2025.04 | 76.4 | |
| GraphMAEBackbone=GraphMAE, Training=Pre-trained2025.04 | 76.3 | |
| GraphLoG2025.04 | 76 | |
| GINE+GPSEBackbone=GINE, SE=GPSE2025.04 | 75.8 | |
| GCCBackbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 75.5 | |
| No pretrainBackbone=GIN, Evaluation Protocol=Fine-tuning2021.10 | 75.4 | |
| GINEBackbone=GINE, Training=from-scratch2025.04 | 74.8 | |
| N-GramClassifier=XGB2026.04 | 74.8 | |
| GraphCL2025.04 | 74.5 | |
| GPS+GFSEBackbone=GPS, SE=GFSE2025.04 | 73.6 | |
| GPF-plusPre-training Strategy=GCL2022.09 | 72.94 | |
| SUPTsoftPre-training Strategy=GCL2024.02 | 70.92 | |
| GPFPre-training Strategy=GCL2024.02 | 70.81 | |
| SUPThardPre-training Strategy=GCL2024.02 | 70.64 | |
| GPF-plusPre-training Strategy=GCL2024.02 | 70.54 | |
| GPS+LapPEBackbone=GPS, SE=LapPE2025.04 | 70.1 | |
| GPFPre-training Strategy=GCL2022.09 | 70.09 | |
| FTPre-training Strategy=GCL2022.09 | 69.78 | |
| FTPre-training Strategy=GCL2024.02 | 69.78 | |
| GPS+RWSEBackbone=GPS, SE=RWSE2025.04 | 69.7 | |
| EGNN2026.04 | 68.6 | |
| GPS+GPSEBackbone=GPS, SE=GPSE2025.04 | 68.3 | |
| SchNet2026.04 | 68.2 | |
| GPSBackbone=GPS, Training=from-scratch2025.04 | 68 | |
| Mol-GDL2026.04 | 67.5 | |
| GROVEModel Variant=base2026.04 | 67.3 | |
| GROVEModel Variant=large2026.04 | 67.3 | |
| GPPTPre-training Strategy=EdgePred2022.09 | 63.05 | |
| GPPTPre-training Strategy=EdgePred2024.02 | 63.05 | |
| GraphPromptPre-training Strategy=EdgePred2022.09 | 62.35 | |
| GraphPromptPre-training Strategy=EdgePred2024.02 | 62.35 |