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

The rise of powerful large language models (LLMs) brings about tremendousopportunities for innovation but also looming risks for individuals and societyat large. We have reached a pivotal moment for ensuring that LLMs andLLM-infused applications are developed and deployed responsibly. However, acentral pillar of responsible AI -- transparency -- is largely missing from thecurrent discourse around LLMs. It is paramount to pursue new approaches toprovide transparency for LLMs, and years of research at the intersection of AIand human-computer interaction (HCI) highlight that we must do so with ahuman-centered perspective: Transparency is fundamentally about supportingappropriate human understanding, and this understanding is sought by differentstakeholders with different goals in different contexts. In this new era ofLLMs, we must develop and design approaches to transparency by considering theneeds of stakeholders in the emerging LLM ecosystem, the novel types ofLLM-infused applications being built, and the new usage patterns and challengesaround LLMs, all while building on lessons learned about how people process,interact with, and make use of information. We reflect on the unique challengesthat arise in providing transparency for LLMs, along with lessons learned fromHCI and responsible AI research that has taken a human-centered perspective onAI transparency. We then lay out four common approaches that the community hastaken to achieve transparency -- model reporting, publishing evaluationresults, providing explanations, and communicating uncertainty -- and call outopen questions around how these approaches may or may not be applied to LLMs.We hope this provides a starting point for discussion and a useful roadmap forfuture research.

AI Transparency in the Age of LLMs: A Human-Centered Research Roadmap