List your company
ResearchFormerly WizWand

Graph Regression on OGB-LSC PCQM4M v2 (val)

0.0809MAEGPTrans-L
0.078620.094010.10940.12479Aug 7, 2021Jan 5, 2022Jun 6, 2022Nov 5, 2022Apr 5, 2023Sep 4, 2023Feb 3, 2024
Updated 8mo ago

Evaluation results

MethodLinks
GPTrans-L#Param=86.0M2023.050.0809
GPTrans-B#Param=45.7M2023.050.0813
GPTrans-S#Param=13.6M2023.050.0823
TIGT# Param=13.0M2024.020.0826
GPTrans-T#Param=6.6M2023.050.0833
GPS-Deep#Param=138.1M2023.050.0852
CSA-deepModel Class=Transformers, # Param.=8.3M2023.040.0853
GPS-medium# Param.=19.4M2022.050.0858
GPS-mediumModel Class=Transformers, # Param.=19.4M2023.040.0858
GPS-mediumMethod Family=Graph Transformers, # Param=19.4M2023.050.0858
GPS-Medium#Param=19.4M2023.050.0858
GPS-medium# Param=19.4M2024.020.0858
GRITMethod Family=Graph Transformers, # Param=16.6M2023.050.0859
EGT-Larger#Param=110.8M2023.050.0859
GRIT# Param=16.6M2024.020.0859
Graphormer + GFSA#Params=48.3M2023.120.086
Graphormer#Params=48.3M2023.120.0862
Graphormer# Param.=48.3M2022.050.0864
GraphormerModel Class=Transformers, # Param.=48.3M2023.040.0864
GraphormerMethod Family=Graph Transformers, # Param=48.3M2023.050.0864
Graphormer# Param=48.3M2024.020.0864
GRPE-Large#Param=118.3M2023.050.0867
EGT-Large (24 layers)#Param=89.3M, Layers=242021.080.0869
EGT# Param.=89.3M2022.050.0869
EGTModel Class=Transformers, # Param.=89.3M2023.040.0869
EGT-Large#Param=89.3M2023.050.0869
EGT-Medium (18 layers)#Param=47.4M, Layers=182021.080.0881
EGT-Medium#Param=47.4M2023.050.0881
GraphormerV2-L#Param=159.3M2023.050.0883
GRPE# Param.=46.2M2022.050.089
GRPEModel Class=Transformers, # Param.=46.2M2023.040.089
GRPEMethod Family=Graph Transformers, # Param=46.2M2023.050.089
GRPE-Standard#Param=46.2M2023.050.089
GRPE# Param=46.2M2024.020.089
CSA-smallModel Class=Transformers, # Param.=2.8M2023.040.0898
EGT-Small (6 layers)#Param=11.5M, Layers=62021.080.0899
EGT-Small#Param=11.5M2023.050.0899
Graphormer#Param=47.1M2021.080.0906
Graphormer-B#Param=47.1M2023.050.0906
TokenGT (Lap)Method Family=Graph Transformers, # Param=48.5M2023.050.091
Graphormer-S#Param=12.5M2023.050.091
TokenGT#Param=48.5M2023.050.091
TokenGTencoding=Lap, # Param=48.5M2024.020.091
GPS-small# Param.=6.2M2022.050.0938
GPS-smallModel Class=Transformers, # Param.=6.2M2023.040.0938
GPS-smallMethod Family=Graph Transformers, # Param=6.2M2023.050.0938
GPS-Small#Param=6.2M2023.050.0938
Cy2C-GNNs# Param=4M2024.020.0956
TokenGT (ORF)Method Family=Graph Transformers, # Param=48.6M2023.050.0962
SEM-ILBackbone=GIN, Algorithm=SEM-IL2023.100.098
SEM-ILBackbone=GCN, Algorithm=SEM-IL2023.100.102
SEM-onlyBackbone=GCN, Algorithm=SEM-only2023.100.106
SEM-onlyBackbone=GIN, Algorithm=SEM-only2023.100.106
Baseline+Backbone=GIN, Algorithm=Baseline+2023.100.108
GIN-VN#Param=6.7M2021.080.1083
GIN-virtual# Param.=6.7M2022.050.1083
GIN-virtualModel Class=MPNN, # Param.=6.7M2023.040.1083
GIN-virtualMethod Family=MPNNs, # Param=6.7M2023.050.1083
GIN-VN#Param=6.7M2023.050.1083
GIN-virtual# Param=6.7M2024.020.1083
BaselineBackbone=GIN, Algorithm=Baseline2023.100.109
GCN-VN#Param=4.9M2021.080.1153
GCN-virtual# Param.=4.9M2022.050.1153
GCN-virtualModel Class=MPNN, # Param.=4.9M2023.040.1153
GCN-virtualMethod Family=MPNNs, # Param=4.9M2023.050.1153
GCN-VN#Param=4.9M2023.050.1153
Baseline+Backbone=GCN, Algorithm=Baseline+2023.100.118
GIN#Param=3.8M2021.080.1195
GIN# Param.=3.8M2022.050.1195
GINModel Class=MPNN, # Param.=3.8M2023.040.1195
GINMethod Family=MPNNs, # Param=3.8M2023.050.1195
GIN#Param=3.8M2023.050.1195
GIN# Param=3.8M2024.020.1195
GCN-virtual# Param=4.9M2024.020.1195
BaselineBackbone=GCN, Algorithm=Baseline2023.100.125
GCN#Param=2.0M2021.080.1379
GCN# Param.=2.0M2022.050.1379
GCNModel Class=MPNN, # Param.=2.0M2023.040.1379
GCNMethod Family=MPNNs, # Param=2.0M2023.050.1379
GCN#Param=2.0M2023.050.1379
GCN# Param=2.0M2024.020.1379