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

Transfer learning through large pre-trained models has changed the landscapeof current applications in natural language processing (NLP). Recently Optimus,a variational autoencoder (VAE) which combines two pre-trained models, BERT andGPT-2, has been released, and its combination with generative adversialnetworks (GANs) has been shown to produce novel, yet very human-looking text.The Optimus and GANs combination avoids the troublesome application of GANs tothe discrete domain of text, and prevents the exposure bias of standard maximumlikelihood methods. We combine the training of GANs in the latent space, withthe finetuning of the decoder of Optimus for single word generation. Thisapproach lets us model both the high-level features of the sentences, and thelow-level word-by-word generation. We finetune using reinforcement learning(RL) by exploiting the structure of GPT-2 and by adding entropy-basedintrinsically motivated rewards to balance between quality and diversity. Webenchmark the results of the VAE-GAN model, and show the improvements broughtby our RL finetuning on three widely used datasets for text generation, withresults that greatly surpass the current state-of-the-art for the quality ofthe generated texts.

OptAGAN: Entropy-based finetuning on text VAE-GAN