Shortcut Learning of Large Language Models in Natural Language Understanding

Mengnan Du; Fengxiang He; Na Zou; Dacheng Tao; Xia Hu

doi:10.1145/3596490

Shortcut Learning of Large Language Models in Natural Language Understanding

Mengnan Du, Fengxiang He, Na Zou, Dacheng Tao, Xia Hu

Data Science

Research output: Contribution to journal › Article › peer-review

58 Scopus citations

Abstract

Shortcuts often hinder the robustness of large language models.

Original language	English (US)
Pages (from-to)	110-120
Number of pages	11
Journal	Communications of the ACM
Volume	67
Issue number	1
DOIs	https://doi.org/10.1145/3596490
State	Published - Dec 21 2023

All Science Journal Classification (ASJC) codes

General Computer Science

Access to Document

10.1145/3596490

Cite this

@article{f031684de1704deab78a5da1b6d2c795,

title = "Shortcut Learning of Large Language Models in Natural Language Understanding",

abstract = "Shortcuts often hinder the robustness of large language models.",

author = "Mengnan Du and Fengxiang He and Na Zou and Dacheng Tao and Xia Hu",

note = "Publisher Copyright: {\textcopyright} 2023 ACM.",

year = "2023",

month = dec,

day = "21",

doi = "10.1145/3596490",

language = "English (US)",

volume = "67",

pages = "110--120",

journal = "Communications of the ACM",

issn = "0001-0782",

publisher = "Association for Computing Machinery (ACM)",

number = "1",

}

TY - JOUR

T1 - Shortcut Learning of Large Language Models in Natural Language Understanding

AU - Du, Mengnan

AU - He, Fengxiang

AU - Zou, Na

AU - Tao, Dacheng

AU - Hu, Xia

PY - 2023/12/21

Y1 - 2023/12/21

N2 - Shortcuts often hinder the robustness of large language models.

AB - Shortcuts often hinder the robustness of large language models.

UR - https://www.scopus.com/pages/publications/85179985005

UR - https://www.scopus.com/inward/citedby.url?scp=85179985005&partnerID=8YFLogxK

U2 - 10.1145/3596490

DO - 10.1145/3596490

M3 - Article

AN - SCOPUS:85179985005

SN - 0001-0782

VL - 67

SP - 110

EP - 120

JO - Communications of the ACM

JF - Communications of the ACM

IS - 1

ER -

Shortcut Learning of Large Language Models in Natural Language Understanding

Abstract

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this