Graph Regression On Zinc

MAE

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	MAE
recipe-for-a-general-powerful-scalable-graph	0.070 ± 0.002
extending-the-design-space-of-graph-neural-1	0.059
generalizing-topological-graph-neural	0.096
towards-better-graph-representation-learning	0.066 ± 0.002
masked-attention-is-all-you-need-for-graphs	0.051
self-attention-in-colors-another-take-on	0.056
graph-learning-with-1d-convolutions-on-random	0.088
a-generalization-of-vit-mlp-mixer-to-graphs	0.075 ± 0.001
from-primes-to-paths-enabling-fast-multi	0.297
multi-mask-aggregators-for-graph-neural	0.156
graph-inductive-biases-in-transformers	0.059
an-experimental-study-of-the-transferability	0.360
learning-long-range-dependencies-on-graphs	0.065 ± 0.001
factorizable-graph-convolutional-networks	0.366
substructure-aware-graph-neural-networks	0.072±0.002
weisfeiler-and-lehman-go-cellular-cw-networks	0.094
principal-neighbourhood-aggregation-for-graph	0.142
graph-transformers-without-positional	0.077
cin-enhancing-topological-message-passing	0.074
topology-informed-graph-transformer	0.057
recipe-for-a-general-powerful-scalable-graph	0.070 ± 0.004
graph-learning-with-1d-convolutions-on-random	0.101
graph-level-representation-learning-with	0.434
ckgconv-general-graph-convolution-with	0.059
cin-enhancing-topological-message-passing	0.091
cin-enhancing-topological-message-passing	0.077
weisfeiler-and-lehman-go-cellular-cw-networks	0.079