Deep equilibrium models as estimators for continuous latent variables
This work provides foundational principles for unsupervised DEQs, potentially benefiting researchers in machine learning and statistics by offering interpretable connections between neural architectures and statistical models.
The paper tackles the problem of estimating continuous latent variables in a generalized exponential family setting with nonlinear transformations, showing that deep equilibrium models (DEQs) solve MAP estimates for latents and parameters, and linking neural network components like activation functions and dropout to statistical assumptions.
Principal Component Analysis (PCA) and its exponential family extensions have three components: observations, latents and parameters of a linear transformation. We consider a generalised setting where the canonical parameters of the exponential family are a nonlinear transformation of the latents. We show explicit relationships between particular neural network architectures and the corresponding statistical models. We find that deep equilibrium models -- a recently introduced class of implicit neural networks -- solve maximum a-posteriori (MAP) estimates for the latents and parameters of the transformation. Our analysis provides a systematic way to relate activation functions, dropout, and layer structure, to statistical assumptions about the observations, thus providing foundational principles for unsupervised DEQs. For hierarchical latents, individual neurons can be interpreted as nodes in a deep graphical model. Our DEQ feature maps are end-to-end differentiable, enabling fine-tuning for downstream tasks.