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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Due to the prohibitively high cost of creating error correction datasets,most Factual Claim Correction methods rely on a powerful verification model toguide the correction process. This leads to a significant drop in performancein domains like scientific claims, where good verification models do not alwaysexist. In this work, we introduce SciFix, a scientific claim correction systemthat does not require a verifier but can outperform existing methods by aconsiderable margin -- achieving correction accuracy of 84% on the SciFactdataset, 77% on SciFact-Open and 72% on the CovidFact dataset, compared to nextbest accuracies of 7%, 5%, and 15% on the same datasets respectively. Ourmethod leverages the power of prompting with LLMs during training to create arichly annotated dataset that can be used for fully supervised training andregularization. We additionally use a claim-aware decoding procedure to improvethe quality of corrected claims. Our method outperforms the very LLM that wasused to generate the annotated dataset -- with Few-Shot Prompting on GPT3.5achieving 58%, 61%, and 64% on the respective datasets, a consistently lowercorrection accuracy, despite using nearly 800 times as many parameters as ourmodel.

SciFix: Outperforming GPT3 on Scientific Factual Error Correction