Graph Property Prediction on ogbg-code2

Metric: Number of params (higher is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	Number of params▼	Extra Data	Paper	Date↕	Code
1	GMAN+bag of tricks	63684290	No	-	-	-
2	DAGNN	35246814	No	-	-	-
3	SAT	15734000	No	Structure-Aware Transformer for Graph Representa...	2022-02-07	Code
4	DAGformer	14952882	No	Transformers over Directed Acyclic Graphs	2022-10-24	Code
5	SAT++ with Magnetic Laplacian	14378069	No	Transformers Meet Directed Graphs	2023-01-31	Code
6	SAT++ with Magnetic Laplacian	14378069	No	Transformers Meet Directed Graphs	2023-01-31	Code
7	GIN+virtual node	13841815	No	How Powerful are Graph Neural Networks?	2018-10-01	Code
8	GCN+virtual node	12484310	No	Semi-Supervised Classification with Graph Convol...	2016-09-09	Code
9	GPS	12454066	No	Recipe for a General, Powerful, Scalable Graph T...	2022-05-25	Code
10	GIN	12390715	No	How Powerful are Graph Neural Networks?	2018-10-01	Code
11	EGC-S (No Edge Features)	11156530	No	Do We Need Anisotropic Graph Neural Networks?	2021-04-03	Code
12	GCN	11033210	No	Semi-Supervised Classification with Graph Convol...	2016-09-09	Code
13	GAT	11030210	No	Graph Attention Networks	2017-10-30	Code
14	PNA (No Edge Features)	10992050	No	Do We Need Anisotropic Graph Neural Networks?	2021-04-03	Code
15	EGC-M (No Edge Features)	10986002	No	Do We Need Anisotropic Graph Neural Networks?	2021-04-03	Code
16	MPNN-Max (No Edge Features)	10971506	No	Do We Need Anisotropic Graph Neural Networks?	2021-04-03	Code
17	DiffPool w/ graphSAGE	10095826	No	Hierarchical Graph Representation Learning with ...	2018-06-22	Code
18	GraphTrans (GCN-Virtual)	9053246	No	-	-	-
19	GraphTrans (GCN)	7563746	No	-	-	-

#1GMAN+bag of tricks
63684290
Number of params
No paper
#2DAGNN
35246814
Number of params
No paper
#3SATSOTA
15734000
Number of params· 2022-02-07
Structure-Aware Transformer for Graph Representation Learning Code
#4DAGformer
14952882
Number of params· 2022-10-24
Transformers over Directed Acyclic Graphs Code
#5SAT++ with Magnetic Laplacian
14378069
Number of params· 2023-01-31
Transformers Meet Directed Graphs Code
#6SAT++ with Magnetic Laplacian
14378069
Number of params· 2023-01-31
Transformers Meet Directed Graphs Code
#7GIN+virtual nodeSOTA
13841815
Number of params· 2018-10-01
How Powerful are Graph Neural Networks?Code
#8GCN+virtual nodeSOTA
12484310
Number of params· 2016-09-09
Semi-Supervised Classification with Graph Convolutional Networks Code
#9GPS
12454066
Number of params· 2022-05-25
Recipe for a General, Powerful, Scalable Graph Transformer Code
#10GIN
12390715
Number of params· 2018-10-01
How Powerful are Graph Neural Networks?Code
#11EGC-S (No Edge Features)
11156530
Number of params· 2021-04-03
Do We Need Anisotropic Graph Neural Networks?Code
#12GCN
11033210
Number of params· 2016-09-09
Semi-Supervised Classification with Graph Convolutional Networks Code
#13GAT
11030210
Number of params· 2017-10-30
Graph Attention Networks Code
#14PNA (No Edge Features)
10992050
Number of params· 2021-04-03
Do We Need Anisotropic Graph Neural Networks?Code
#15EGC-M (No Edge Features)
10986002
Number of params· 2021-04-03
Do We Need Anisotropic Graph Neural Networks?Code
#16MPNN-Max (No Edge Features)
10971506
Number of params· 2021-04-03
Do We Need Anisotropic Graph Neural Networks?Code
#17DiffPool w/ graphSAGE
10095826
Number of params· 2018-06-22
Hierarchical Graph Representation Learning with Differentiable Pooling Code
#18GraphTrans (GCN-Virtual)
9053246
Number of params
No paper
#19GraphTrans (GCN)
7563746
Number of params
No paper