Flow-Matching Models Outperform VAEs and GANs in Preserving Feature Dependencies for Rare-Class Augmentation

This report synthesises findings from 3 peer-reviewed papers addressing the following research question: How do flow-matching models for tabular data perform relative to VAEs and GANs in preserving feature dependencies when generating synthetic samples for rare-class augmentation, as measured by mutual. 9 claims were extracted from source literature; 9 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.8/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: How do flow-matching models for tabular data perform relative to VAEs and GANs in preserving feature dependencies when generating synthetic samples for rare-class augmentation, as measured by mutual information and classification recall?

Autonomous literature synthesis. Automated review score: 8.8/10. Full text and citation available at Assignee Research.