Efficient and Scalable Graph Generation through Iterative Local Expansion
A novel graph generation method that efficiently generates graphs by iteratively expanding a single node into the target graph through localized denoising diffusion, capturing both global and local graph structures.
A Simple and Scalable Representation for Graph Generation at ICLR 2024
Introducing GEEL for scalable graph generation.
Graph Generation with K2–Trees: A Novel Approach for Hierarchical and Compact Graph Generation
Novel method leveraging K2–tree for hierarchical and compact graph generation.
Exploring Large Language Models for Graph Generation
Large language models show potential in graph generation tasks, including rule-based and distribution-based generation, with preliminary abilities in generating molecules with specific properties.
GraphRCG: Self-conditioned Graph Generation via Bootstrapped Representations
Proposing a novel self-conditioned graph generation framework to explicitly model graph distributions and guide the generation process using bootstrapped representations.
GraphRCG: Self-conditioned Graph Generation Framework
In this work, the authors propose a novel self-conditioned graph generation framework that explicitly models graph distributions and utilizes them to guide the generation process. By capturing graph distributions through representations and employing self-conditioned guidance, the framework enhances the fidelity of generated graphs.
© 2024 by Linnk AI