English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Variational autoencoders (VAEs) suffer from posterior collapse, where thepowerful neural networks used for modeling and inference optimize the objectivewithout meaningfully using the latent representation. We introduce inferencecritics that detect and incentivize against posterior collapse by requiringcorrespondence between latent variables and the observations. By connecting thecritic's objective to the literature in self-supervised contrastiverepresentation learning, we show both theoretically and empirically thatoptimizing inference critics increases the mutual information betweenobservations and latents, mitigating posterior collapse. This approach isstraightforward to implement and requires significantly less training time thanprior methods, yet obtains competitive results on three established datasets.Overall, the approach lays the foundation to bridge the previously disconnectedframeworks of contrastive learning and probabilistic modeling with variationalautoencoders, underscoring the benefits both communities may find at theirintersection.

Forget-me-not! Contrastive Critics for Mitigating Posterior Collapse