figshare
Browse

All experiments conducted on ChatGPT-3.5 and ChatGPT-4, which are the subject of this study, are included in this file.

Download (20.17 MB)
journal contribution
posted on 2024-10-23, 17:22 authored by Marco Giunti, Fabrizia Giulia Garavaglia, Roberto Giuntini, Giuseppe Sergioli, Simone Pinna

All experiments conducted on ChatGPT-3.5 and ChatGPT-4, which are the subject of this study, are included in this file.

History

Usage metrics

    PLOS ONE

    Categories

    Keywords

    three different typesrequire specific knowledgemodels effectively despitecomputational errors madescholastic assessment testreveal important weaknessesuniversity admission testssat testxlink ">two teststop 10significant improvementsat ),real studentsprovides insightprospective undergraduatelooked closelyincorrect answersimsat ).experiment conductedbmat ),biomedical programsaverage withinarticle reports

    Licence

    CC BY 4.0

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC