Outreach

Gave a talk about my recent works on over-squashing, over-smoothing and memorization in Graph Neural Networks.

Had a fun time talking about our recently accepted NeurIPS paper with neptune.ai. The goal was to explain our research using simpler analogies intended for broader audience, making complex graph neural network concepts accessible to a wider community.

Presented our recent work on addressing fundamental challenges in Graph Neural Networks - specifically tackling both over-squashing and over-smoothing phenomena through spectral pruning techniques. The presentation covered theoretical insights and practical applications.