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-context learning (ICL) is a new learning paradigm that has gainedpopularity along with the development of large language models. In this work,we adapt a recently proposed hardness metric, pointwise $\mathcal{V}$-usableinformation (PVI), to an in-context version (in-context PVI). Compared to theoriginal PVI, in-context PVI is more efficient in that it requires only a fewexemplars and does not require fine-tuning. We conducted a comprehensiveempirical analysis to evaluate the reliability of in-context PVI. Our findingsindicate that in-context PVI estimates exhibit similar characteristics to theoriginal PVI. Specific to the in-context setting, we show that in-context PVIestimates remain consistent across different exemplar selections and numbers ofshots. The variance of in-context PVI estimates across different exemplarselections is insignificant, which suggests that in-context PVI are stable.Furthermore, we demonstrate how in-context PVI can be employed to identifychallenging instances. Our work highlights the potential of in-context PVI andprovides new insights into the capabilities of ICL.

Measuring Pointwise $\mathcal{V}$-Usable Information In-Context-ly