Chiral Tag Mislabelling

[RylieWeaver]

Apologies for the slightly misleading issue. I couldn’t find a good email to contact the ChiENN group, so I’m hoping to grab your attention here.

I’m Rylie, a researcher at Oak Ridge National Lab, investigating chirality in GNNs. I found inspiration from your ChiENN paper but am encountering issues with chiral tags that don't match the positions/orderings of neighboring nodes. I’m likely missing something, and have been stuck on this problem for a while, so it would be an enormous help to get some insight on what the issue might be here. I've set up a Google Colab notebook to illustrate the issue, and it’s ready to run out-of-the-box with explanations included: Link.

Thanks,
Rylie Weaver

[email protected]
Oak Ridge National Laboratory | Research Intern
Claremont Graduate University | MS Machine Learning
University of Redlands | BS Mathematics (Honors) and Physics
LinkedIn | Github