On the Effectiveness of Random Weights in Graph Neural Networks
Thu Bui, Carola-Bibiane Schönlieb, Bruno Ribeiro, Beatrice Bevilacqua, Moshe Eliasof · 2025
Read On the Effectiveness of Random Weights in Graph Neural Networks by Thu Bui, Carola-Bibiane Schönlieb, Bruno Ribeiro, Beatrice Bevilacqua, Moshe Eliasof (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-00190.txt.
Description
Thu Bui and co-authors show in 2025 that graph neural networks can work well even when their weights are random rather than learned. Such networks pass messages between linked nodes to learn from graph shaped data.
Swapping learned weights for random ones kept predictive power close to fully trained models across many tasks and data sets. Training time fell by up to six times and memory use by up to three times. The random weights also form random spread operators that fight a known flaw in these networks, the collapse of feature rank. The authors offer random weights as a light, cheap option for graph model design.
Catalog record
| CREATOR | Thu Bui, Carola-Bibiane Schönlieb, Bruno Ribeiro, Beatrice Bevilacqua, Moshe Eliasof |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-00190