A latent-observed dissimilarity measure
This work addresses a methodological gap for researchers in generative models and representation learning, though it appears incremental as it builds on existing frameworks.
The authors tackled the problem of quantitatively assessing relationships between latent and observed variables in generative models by proposing a latent-observed dissimilarity (LOD) measure, and experiments on real-world data showed it effectively captures model differences and reflects higher-layer learning capabilities, with conditional independence improving information transmission.
Quantitatively assessing relationships between latent variables and observed variables is important for understanding and developing generative models and representation learning. In this paper, we propose latent-observed dissimilarity (LOD) to evaluate the dissimilarity between the probabilistic characteristics of latent and observed variables. We also define four essential types of generative models with different independence/conditional independence configurations. Experiments using tractable real-world data show that LOD can effectively capture the differences between models and reflect the capability for higher layer learning. They also show that the conditional independence of latent variables given observed variables contributes to improving the transmission of information and characteristics from lower layers to higher layers.