A semi-amortized method is introduced that divides parameters into smaller blocks, and uses conditional independence to identify the information needed to infer each block, and trains neural estimators for blocks whose posterior updates are computationally expensive or frequently repeated.
Niek Stevenson, Paul-Christian Burkner, Andrew Heathcote et al.· 0 citations
We develop a general method for amortized Bayesian inference on multilevel models of arbitrary structure. Given a generative model specified as a directed acyclic graph, our method automatically derives valid factorizations of the joint posterior and matching neural network architectures. The key steps, graph expansion...
Daniel Habermann, Andreas Bulling, Stefan T. Radev et al.· 0 citations
This work shows how auxiliary guidance losses applied to internal representations leads to faster convergence when training data is abundant and to better performance when it is scarce, and demonstrates that auxiliary supervision improves convergence speed and accuracy on a range of challenging real-world inference pro...
Hans Olischläger, Svenja Jedhoff, Š. Kucharský et al.· arXiv.org· 0 citations
This work describes the conditions under which latent-space inference recovers the desired target posterior and systematically study its empirical trade-offs, comparing latent and standard estimators while controlling for network capacity, regularization, optimization, and training compute.
Lars Kühmichel, Stefan T. Radev, B. Koppolu et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.