Principled Latent Diffusion for Graphs via Laplacian Autoencoders

Antoine Siraudin, Christopher Morris · 2026

Read Principled Latent Diffusion for Graphs via Laplacian Autoencoders by Antoine Siraudin, Christopher Morris (2026) 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-2601-13780.txt.

Description

Antoine Siraudin and Christopher Morris build a faster way to generate new graphs with AI in 2026. Graph diffusion models make state-of-the-art graphs, from molecules to social nets. Their cost grows with the square of the node count. Much of that cost is spent modeling edges that are not there, in sparse graphs.

The pair compress graphs into a low-dimensional latent space first. Diffusion then runs in that small space, not on the full graph. Laplacian autoencoders do the compression and rebuild steps. Latent diffusion has cut costs in images, and may do the same for graphs.

Catalog record

CREATORAntoine Siraudin, Christopher Morris
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE92 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-13780

Browse the full library · About the Tower of Babel