CRaWl

Reported on 5 benchmarks across 4 tasks · 1 paper · 3 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Graphs4 results

Graph RegressiononZINC
MAE· 2021-02-17
0.101
best: 0.051 (ESA + rings + NodeRWSE + EdgeRWSE)
SOTA
Walking Out of the Weisfeiler Leman Hierarchy: Graph Learning Beyond Message Passing arXiv:2102.08786
Graph ClassificationonREDDIT-B
Accuracy· 2021-02-17
93.15
SOTA
Walking Out of the Weisfeiler Leman Hierarchy: Graph Learning Beyond Message Passing arXiv:2102.08786
Graph RegressiononZINC-500k
MAE· 2021-02-17
0.101
best: 0.051 (ESA + rings + NodeRWSE + EdgeRWSE)
Walking Out of the Weisfeiler Leman Hierarchy: Graph Learning Beyond Message Passing arXiv:2102.08786
Graph Property Predictiononogbg-molpcba
Number of params· 2021-02-17
6115728
best: 119529665 (HIG(pre-trained on PCQM4M))
Walking Out of the Weisfeiler Leman Hierarchy: Graph Learning Beyond Message Passing arXiv:2102.08786

Methodology1 result

ClassificationonREDDIT-B
Accuracy· 2021-02-17
93.15
SOTA
Walking Out of the Weisfeiler Leman Hierarchy: Graph Learning Beyond Message Passing arXiv:2102.08786