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

Conditional variational models, using either continuous or discrete latentvariables, are powerful for open-domain dialogue response generation. However,previous works show that continuous latent variables tend to reduce thecoherence of generated responses. In this paper, we also found that discretelatent variables have difficulty capturing more diverse expressions. To tacklethese problems, we combine the merits of both continuous and discrete latentvariables and propose a Hybrid Latent Variable (HLV) method. Specifically, HLVconstrains the global semantics of responses through discrete latent variablesand enriches responses with continuous latent variables. Thus, we diversify thegenerated responses while maintaining relevance and coherence. In addition, wepropose Conditional Hybrid Variational Transformer (CHVT) to construct and toutilize HLV with transformers for dialogue generation. Through fine-grainedsymbolic-level semantic information and additive Gaussian mixing, we constructthe distribution of continuous variables, prompting the generation of diverseexpressions. Meanwhile, to maintain the relevance and coherence, the discretelatent variable is optimized by self-separation training. Experimental resultson two dialogue generation datasets (DailyDialog and Opensubtitles) show thatCHVT is superior to traditional transformer-based variational mechanism w.r.t.diversity, relevance and coherence metrics. Moreover, we also demonstrate thebenefit of applying HLV to fine-tuning two pre-trained dialogue models (PLATOand BART-base).

Towards Diverse, Relevant and Coherent Open-Domain Dialogue Generation via Hybrid Latent Variables