cs.LGFeb 5, 2026

Disentangled Representation Learning via Flow Matching

Authors: Jinjin Chi, Taoping Liu, Mengtao Yin, Ximing Li, Yongcheng Jing, Jialie Shen, Leszek Rutkowski, Dacheng Tao

Organizations: College of Computer Science and Technology, Jilin University, Changchun, China · College of Computing and Data Science, Nanyang Technological University, Singapore · City St George’s, University of London, London, United Kingdom · Systems Research Institute, Polish Academy of Sciences, Warsaw 01-447, Poland

Abstract

Disentangled representation learning aims to capture the underlying explanatory factors of observed data, enabling a principled understanding of the data-generating process. Recent advances in generative modeling have introduced new paradigms for learning such representations. However, existing diffusion-based methods encourage factor independence via inductive biases, yet frequently lack strong semantic alignment. In this work, we propose a flow matching-based framework for disentangled representation learning, which casts disentanglement as learning factor-conditioned flows in a compact latent space. To enforce explicit semantic alignment, we introduce a non-overlap (orthogonality) regularizer that suppresses cross-factor interference and reduces information leakage between factors. Extensive experiments across multiple datasets demonstrate consistent improvements over representative baselines, yielding higher disentanglement scores as well as improved controllability and sample fidelity.

Figures & tables

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Unsupervised Disentanglement Without Compromises : How Functional Orthogonality Enforces Identifiability

    Jun 19, 2026Mathieu Cyrille Simon, Pascal Frossard, Christophe De VleeschouwerUnsupervised LearningDisentangled Representation Learning

  2. The Trade-off Between Covariate Dependence and Latent Structure in Representation Learning

    Aug 17, 2026Małgorzata Łazęcka, Ewa SzczurekDisentangled Representation LearningLatent Space Alignment