Efficient Graph Generation via Direct Prediction and Flow Matching

arXiv:2610.05397 · cs.LG · Submitted 2026-10-04 · Read on arXiv

cs.LG

Submitted: 2026-10-04

Updated: 2026-10-04

Code: https://github.com/Susie-Lu/DiGFM

Terminology

• Flow Matching — 1× in this paper · explained in 10 episodes

Sources

• Non-Denoising Forward-Time Diffusions
• Principled Latent Diffusion for Graphs via Laplacian Autoencoders
• Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Design
• Categorical Normalizing Flows via Continuous Transformations
• Cometh: A continuous-time discrete-state graph diffusion model

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks