Multiresolution Graph Transformers and Wavelet Positional Encoding for Learning Hierarchical Structures

Open paper

Benchmark	Model	Rank	Results
graph-classification-on-peptides-func	MGT+WavePE	#21	AP: 0.6817±0.0064
graph-regression-on-peptides-struct	MGT+WavePE	#7	MAE: 0.2453±0.0025