Reproduce “Improved Generalization of Weight Space Networks via Augmentations” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Aviv Shamsian, Aviv Navon, David W. Zhang, Yan Zhang, Ethan Fetaya, Gal Chechik, Haggai Maron
The authors introduce alignment-based MixUp to augment weight-space learning, where neural network weights encode objects such as 3D shapes. They approximately match hidden-neuron orderings in two networks, then blend their weights and labels. On one-view ModelNet40 classification, a GNN reaches 26.88 ± 1.75% accuracy, versus 8.82 ± 1.08% without augmentation; evidence is limited to tested INR settings.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.