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

Multi-aspect controllable text generation aims to generate fluent sentencesthat possess multiple desired attributes simultaneously. Traditional methodseither combine many operators in the decoding stage, often with costlyiteration or search in the discrete text space, or train separate controllersfor each aspect, resulting in a degeneration of text quality due to thediscrepancy between different aspects. To address these limitations, weintroduce a novel approach for multi-aspect control, namely MacLaSa, thatestimates compact latent space for multiple aspects and performs efficientsampling with a robust sampler based on ordinary differential equations (ODEs).To eliminate the domain gaps between different aspects, we utilize aVariational Autoencoder (VAE) network to map text sequences from varying datasources into close latent representations. The estimated latent space enablesthe formulation of joint energy-based models (EBMs) and the plugging in ofarbitrary attribute discriminators to achieve multi-aspect control. Afterwards,we draw latent vector samples with an ODE-based sampler and feed sampledexamples to the VAE decoder to produce target text sequences. Experimentalresults demonstrate that MacLaSa outperforms several strong baselines onattribute relevance and textual quality while maintaining a high inferencespeed.

MacLaSa: Multi-Aspect Controllable Text Generation via Efficient Sampling from Compact Latent Space