Integrated SSMA aggregation

[AlmogDavid]

Integrated the "Sequential Signal Mixing Aggregation" (SSMA) aggregation method from: https://arxiv.org/abs/2409.19414

[Kh4L]

Looks great! Did you try it out with some OGB benchmarks from PyG examples?

I think it'd be nice to add an option to try it out in ogbn_train.py - in a followup PR, and see if we get the accuracy boost claimed in Table 2 ( like the 15-20% boost in ogbn-products 🚀 )!

[AlmogDavid]

Looks great! Did you try it out with some OGB benchmarks from PyG examples?

I think it'd be nice to add an option to try it out in ogbn_train.py - in a followup PR, and see if we get the accuracy boost claimed in Table 2 ( like the 15-20% boost in ogbn-products 🚀 )!

Thanks!
You can find the experimental setting and to reproduce the results using our repository: https://almogdavid.github.io/SSMA/
There you can find full set of datasets and configurations to try out yourself, not sure if its appropriate to link to the repository from the documentation itself, but if you think it will be helpful I will do it.