Causally Disentangled Generative Variational AutoEncoder
This addresses the challenge of causal disentanglement in generative models for researchers in machine learning, though it appears incremental as it builds on existing VAE frameworks.
The paper tackles the problem of learning causally disentangled representations and generating disentangled outcomes simultaneously in Variational AutoEncoders, introducing a new supervised technique called Causally Disentangled Generation (CDG) and showing that encoder regularization alone is insufficient, with empirical support from image and tabular datasets.
We present a new supervised learning technique for the Variational AutoEncoder (VAE) that allows it to learn a causally disentangled representation and generate causally disentangled outcomes simultaneously. We call this approach Causally Disentangled Generation (CDG). CDG is a generative model that accurately decodes an output based on a causally disentangled representation. Our research demonstrates that adding supervised regularization to the encoder alone is insufficient for achieving a generative model with CDG, even for a simple task. Therefore, we explore the necessary and sufficient conditions for achieving CDG within a specific model. Additionally, we introduce a universal metric for evaluating the causal disentanglement of a generative model. Empirical results from both image and tabular datasets support our findings.