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 Autoencoder (VAE) is widely used as a generative model toapproximate a model's posterior on latent variables by combining the amortizedvariational inference and deep neural networks. However, when paired withstrong autoregressive decoders, VAE often converges to a degenerated localoptimum known as "posterior collapse". Previous approaches consider theKullback Leibler divergence (KL) individual for each datapoint. We propose tolet the KL follow a distribution across the whole dataset, and analyze that itis sufficient to prevent posterior collapse by keeping the expectation of theKL's distribution positive. Then we propose Batch Normalized-VAE (BN-VAE), asimple but effective approach to set a lower bound of the expectation byregularizing the distribution of the approximate posterior's parameters.Without introducing any new model component or modifying the objective, ourapproach can avoid the posterior collapse effectively and efficiently. Wefurther show that the proposed BN-VAE can be extended to conditional VAE(CVAE). Empirically, our approach surpasses strong autoregressive baselines onlanguage modeling, text classification and dialogue generation, and rivals morecomplex approaches while keeping almost the same training time as VAE.

A Batch Normalized Inference Network Keeps the KL Vanishing Away