Reply to Hu et al.: Applying different evaluation standards to humans vs. Large Language Models overestimates AI performance-Reference-Cited by-同舟云学术

登录注册会员服务联系我们

Reply to Hu et al.: Applying different evaluation standards to humans vs. Large Language Models overestimates AI performance

Published:2024-08-26 Issue:36 Volume:121 Page:
ISSN:0027-8424
Container-title:Proceedings of the National Academy of Sciences
language:en
Short-container-title:Proc. Natl. Acad. Sci. U.S.A.

Author:

Leivada Evelina¹^ORCID,Günther Fritz²,Dentella Vittoria³^ORCID

Affiliation:

1. Departament de Filologia Catalana, Universitat Autònoma de Barcelona, Barcelona 08193, Spain

2. Institut für Psychologie, Humboldt Universität zu Berlin, Berlin 10099, Germany

3. Departament d'Estudis Anglesos i Alemanys, Universitat Rovira i Virgili, Tarragona 43002, Spain

Funder

EC | Horizon 2020 Framework Programme

Universitat Rovira i Virgili

Deutsche Forschungsgemeinschaft

Ministerio de Ciencia e Innovación

Publisher

Proceedings of the National Academy of Sciences

Link

https://pnas.org/doi/pdf/10.1073/pnas.2406752121

Reference5 articles.

1. Systematic testing of three Language Models reveals low language accuracy, absence of response stability, and a yes-response bias

2. Language models align with human judgments on key grammatical constructions;Hu J.;Proc. Natl. Acad. Sci. U.S.A.,2024

3. Acceptable Ungrammatical Sentences, Unacceptable Grammatical Sentences, and the Role of the Cognitive Parser

4. Prompting is not a substitute for probability measurements in large language models

5. Evaluating the language abilities of Large Language Models vs. humans: Three caveats

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线，采集、加工和组织学术论文而形成的新型学术文献查询和分析系统，可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容，当前同舟云学术共收录了国内外主流学术期刊6万余种，收集的期刊论文及会议论文总量共计约1.5亿篇，并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询！咨询电话：010-8811{复制后删除}0370

www.globalauthorid.com

TOP