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

A human decision-maker benefits the most from an AI assistant that correctsfor their biases. For problems such as generating interpretation of a radiologyreport given findings, a system predicting only highly likely outcomes may beless useful, where such outcomes are already obvious to the user. To alleviatebiases in human decision-making, it is worth considering a broad differentialdiagnosis, going beyond the most likely options. We introduce a new task, "lesslikely brainstorming," that asks a model to generate outputs that humans thinkare relevant but less likely to happen. We explore the task in two settings: abrain MRI interpretation generation setting and an everyday commonsensereasoning setting. We found that a baseline approach of training with lesslikely hypotheses as targets generates outputs that humans evaluate as eitherlikely or irrelevant nearly half of the time; standard MLE training is noteffective. To tackle this problem, we propose a controlled text generationmethod that uses a novel contrastive learning strategy to encourage models todifferentiate between generating likely and less likely outputs according tohumans. We compare our method with several state-of-the-art controlled textgeneration models via automatic and human evaluations and show that our models'capability of generating less likely outputs is improved.

Less Likely Brainstorming: Using Language Models to Generate Alternative Hypotheses