Logic-Guided Graph Denoising Diffusion Models

Key points

• Graph diffusion models are competitive approaches for generating complex graph-structured data.1
• We incorporate domain knowledge using classifier guidance, a technique that conditions the diffusion model's distribution on external variables to exert control over the sampled data.1
• During the generative process, classifier guidance steers the denoising trajectory to optimize an objective function, which we define as a fuzzy and differentiable relaxation of the logical rules.1
• We evaluate our approach by comparing the unconditional generation of knowledge graphs with our logic-guided framework, which incorporates domain ontologies as structural constraints during inference.1

Sources (1)

• [1]Logic-Guided Graph Denoising Diffusion Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 05:06 PM

  In this paper, we propose a method for grounding graph diffusion models in domain-specific knowledge via First Order Logic (FOL), leveraging its expressive and interpretable formalism to impose structural constraints on the generative process.

  Graph diffusion models are competitive approaches for generating complex graph-structured data.

Extractive summary: sentences quoted from the sources.