OpenCodePapers

graph-property-prediction-on-ogbg-molhiv

Graph Property Prediction
Dataset Link
Results over time
Click legend items to toggle metrics. Hover points for model names.
Leaderboard
Click a sortable column header to sort.
PaperCodeTest ROC-AUCExt. dataValidation ROC-AUCNumber of paramsModelNameReleaseDate
[]()0.8475 ± 0.0003No0.8275 ± 0.00085908027HyperFusion
[]()0.8420 ± 0.0015No0.8238 ± 0.002826706953PAS+FPs
[]()0.8403 ± 0.0021No0.8176 ± 0.00341019408HIG
Large-scale Robust Deep AUC Maximization: A New Surrogate Loss and Empirical Studies on Medical Image Classification✓ Link0.8352 ± 0.0054No0.8238 ± 0.00613444509DeepAUC2020-12-06
[]()0.8244 ± 0.0033No0.8329 ± 0.00391444110FingerPrint+GMAN
Molecular Representation Learning by Leveraging Chemical Information✓ Link0.8232 ± 0.0047No0.8331 ± 0.00542425102Neural FingerPrints2021-03-15
Do Transformers Really Perform Bad for Graph Representation?✓ Link0.8225 ± 0.0001No0.8396 ± 0.000147085378Graphormer + FPs2021-06-09
[]()0.8208 ± 0.0037No0.8036 ± 0.00595782Molecular FP + Random Forest
Graph Propagation Transformer for Graph Representation Learning✓ Link0.8126 ± 0.0032YesGPTrans-B2023-05-19
Weisfeiler and Lehman Go Cellular: CW Networks✓ Link0.8094 ± 0.0057No0.8277 ± 0.0099239745CIN2021-06-23
Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism✓ Link0.8067 ± 0.0950No0.8347 ± 0.0031249602GSAT2022-01-31
[]()0.8060 ± 0.0010No0.8420 ± 0.0030230000MorganFP+Rand. Forest
Global Self-Attention as a Replacement for Graph Convolution✓ Link0.806 ± 0.0065EGT2021-08-07
Weisfeiler and Lehman Go Cellular: CW Networks✓ Link0.8055 ± 0.0104No0.8310 ± 0.0102138337CIN-small2021-06-23
Do Transformers Really Perform Bad for Graph Representation?✓ Link0.8051 ± 0.0053Yes0.8310 ± 0.008947183040Graphormer2021-06-09
Do Transformers Really Perform Bad for Graph Representation?✓ Link0.8051 ± 0.0053Yes0.8310 ± 0.008947183040Graphormer (pre-trained on PCQM4M)2021-06-09
Unlocking the Potential of Classic GNNs for Graph-level Tasks: Simple Architectures Meet Excellence✓ Link0.8040 ± 0.0164No0.8329 ± 0.01581076633GatedGCN+2025-02-13
Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting✓ Link0.8039 ± 0.0090No0.8473 ± 0.0096114211directional GSN2020-06-16
A Persistent Weisfeiler–Lehman Procedure for Graph Classification✓ Link0.8039 ± 0.0040No0.8279 ± 0.00594600000P-WL2019-06-09
Nested Graph Neural Networks✓ Link0.7986 ± 0.01050.8080 ± 0.0278Nested GIN+virtual node (ens)2021-10-25
Directional Graph Networks✓ Link0.7970 ± 0.0097No0.8470 ± 0.0047114065DGN2020-10-06
Weisfeiler and Lehman Go Paths: Learning Topological Features via Path Complexes0.7944 ± 1.40 PIN2023-08-13
Robust Optimization as Data Augmentation for Large-scale Graphs✓ Link0.7942 ± 0.0120No0.8425 ± 0.0061531976DeeperGCN+FLAG2020-10-19
Parameterized Hypercomplex Graph Neural Networks for Graph Classification✓ Link0.7934 ± 0.0116No0.8217 ± 0.0089110909PHC-GNN2021-03-30
Principal Neighbourhood Aggregation for Graph Nets✓ Link0.7905 ± 0.0132No0.8519 ± 0.0099326081PNA2020-04-12
GraphNorm: A Principled Approach to Accelerating Graph Neural Network Training✓ Link0.7883 ± 0.0100No0.7904 ± 0.0115526201GCN+GraphNorm2020-09-07
Recipe for a General, Powerful, Scalable Graph Transformer✓ Link0.7880No0.8255 ± 0.0092558625GPS2022-05-25
Hierarchical Inter-Message Passing for Learning on Molecular Graphs✓ Link0.7880 ± 0.0082NoPlease tell us153029HIMP2020-06-22
DeeperGCN: All You Need to Train Deeper GCNs✓ Link0.7858 ± 0.0117No0.8427 ± 0.0063531976DeeperGCN2020-06-13
Nested Graph Neural Networks✓ Link0.7834 ± 0.01860.8317 ± 0.0199Nested GIN+virtual node2021-10-25
[]()0.7825 ± 0.0121No0.8009 ± 0.007832385GIN
Do We Need Anisotropic Graph Neural Networks?✓ Link0.7818 ± 0.0153No0.8396 ± 0.0097317265EGC-M (No Edge Features)2021-04-03
Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting✓ Link0.7799 ± 0.0100No0.8658 ± 0.00843338701GSN2020-06-16
Wasserstein Embedding for Graph Learning✓ Link0.7757 ± 0.0111No0.8101 ± 0.0097361064WEGL2020-06-16
Robust Optimization as Data Augmentation for Large-scale Graphs✓ Link0.7748 ± 0.0096No0.8438 ± 0.01283336306GIN+virtual node+FLAG2020-10-19
Do We Need Anisotropic Graph Neural Networks?✓ Link0.7721 ± 0.0110No0.8366 ± 0.0074317013EGC-S (No Edge Features)2021-04-03
How Powerful are Graph Neural Networks?✓ Link0.7707 ± 0.0149No0.8479 ± 0.00683336306GIN+virtual node2018-10-01
Robust Optimization as Data Augmentation for Large-scale Graphs✓ Link0.7683 ± 0.0102No0.8176 ± 0.0087527701GCN+FLAG2020-10-19
Robust Optimization as Data Augmentation for Large-scale Graphs✓ Link0.7654 ± 0.0114No0.8225 ± 0.01551885206GIN+FLAG2020-10-19
Semi-Supervised Classification with Graph Convolutional Networks✓ Link0.7606 ± 0.0097No0.8204 ± 0.0141527701GCN2016-09-09
Semi-Supervised Classification with Graph Convolutional Networks✓ Link0.7599 ± 0.0119No0.8384 ± 0.00911978801GCN+virtual node2016-09-09
How Powerful are Graph Neural Networks?✓ Link0.7558 ± 0.0140No0.8232 ± 0.00901885206GIN2018-10-01
Semi-Supervised Classification with Graph Convolutional Networks✓ Link0.7549 ± 0.0163No0.8042 ± 0.0107527701GCN (in Julia)2016-09-09
Check out how to contribute new results.