Learning Causally Disentangled Representations via the Principle of Independent Causal Mechanisms

arXiv:2306.01213v3 Announce Type: replace
Abstract: Learning disentangled causal representations is a challenging problem that has gained significant attention recently due to its implications for extracting meaningful information for downstream tasks. In this work, we define a new notion of causal disentanglement from the perspective of independent causal mechanisms. We propose ICM-VAE, a framework for learning causally disentangled representations supervised by causally related observed labels. We model causal mechanisms using nonlinear learnable flow-based diffeomorphic functions to map noise variables to latent …

abstract arxiv attention causal cs.lg independent information notion perspective stat.ml tasks type via work