Spectral Neural Graph Sparsification
Angelica Liguori, Ettore Ritacco, Pietro Sabatino, Annalisa Socievole · 2025
Read Spectral Neural Graph Sparsification by Angelica Liguori, Ettore Ritacco, Pietro Sabatino, Annalisa Socievole (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-2510-27474.txt.
Description
Angelica Liguori, Ettore Ritacco, and two other co-authors shrink big networks while keeping their shape, in a 2025 paper. Graphs model linked systems: friends, molecules, brain regions. Graph neural networks learn from them, but they lean on a fixed structure and can blur node traits together. A smaller proxy graph would cut costs for tasks like grouping or spread tracing.
Their Spectral Preservation Network grows the smaller graph layer by layer. A joint evolution layer reshapes both the links and the node traits at once, so neither is frozen. A spectral concordance loss, a penalty rule, keeps the small graph's wave-like traits close to the original's. Node traits are kept consistent too. The authors test the proxies on node-level thinning tasks against standard benchmarks.
Catalog record
| CREATOR | Angelica Liguori, Ettore Ritacco, Pietro Sabatino, Annalisa Socievole |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-27474