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

In open-domain dialogues, predictive uncertainties are mainly evaluated in adomain shift setting to cope with out-of-distribution inputs. However, inreal-world conversations, there could be more extensive distributional shiftedinputs than the out-of-distribution. To evaluate this, we first propose twomethods, Unknown Word (UW) and Insufficient Context (IC), enabling gradualdistributional shifts by corruption on the dialogue dataset. We theninvestigate the effect of distributional shifts on accuracy and calibration.Our experiments show that the performance of existing uncertainty estimationmethods consistently degrades with intensifying the shift. The results suggestthat the proposed methods could be useful for evaluating the calibration ofdialogue systems under distributional shifts.

Evaluating Predictive Uncertainty under Distributional Shift on Dialogue Dataset