Research map / Graph learning / Graph self-supervised learning
Beyond Node Sequences: Relational Diffusion for Unified Graph Learning
Xianan Wang, Wenji Hu, Chunyu Wei, Yueguo Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes relational diffusion, a graph learning framework that models graphs as relations via diffusion rather than serializing nodes into token sequences, aiming for a unified graph foundation model.
Problem
Graph foundation models serialize graphs into node token sequences with autoregressive or masked prediction, which disrupts the relation-centric nature of graphs and violates permutation equivariance.
Approach
Reconceptualizes graph modeling as relational diffusion instead of node-sequential modeling, so that the generative process operates over relations and preserves permutation equivariance.
Closest papers by meaning
- ARDiff: Anisotropic Residual Diffusion for Heterogeneous Graph Learning (AAAI 2026 · similarity 0.91)
- Generative Modelling of Structurally Constrained Graphs (NeurIPS 2024 · similarity 0.88)
- Unifying Generation and Prediction on Graphs with Latent Graph Diffusion (NeurIPS 2024 · similarity 0.88)
- A Generalized Neural Diffusion Framework on Graphs (AAAI 2024 · similarity 0.88)
- When Edge Independence Fails: Joint Graph Diffusion with Latent Sociability Priors (NeurIPS 2026 · similarity 0.88)
- Cross-Domain Graph Data Scaling: A Showcase with Diffusion Models (NeurIPS 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
