Using Natural Sentences for Understanding Biases in Language Models

被引:0
|
作者
Alnegheimish, Sarah [1 ]
Guo, Alicia [1 ]
Sun, Yi [1 ]
机构
[1] MIT, Cambridge, MA 02139 USA
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Evaluation of biases in language models is often limited to synthetically generated datasets. This dependence traces back to the need of prompt-style dataset to trigger specific behaviors of language models. In this paper, we address this gap by creating a prompt dataset with respect to occupations collected from real-world natural sentences present in Wikipedia. We aim to understand the differences between using template-based prompts and natural sentence prompts when studying gender-occupation biases in language models. We find bias evaluations are very sensitive to the design choices of template prompts, and we propose using natural sentence prompts for systematic evaluations to step away from design choices that could introduce bias in the observations.
引用
收藏
页码:2824 / 2830
页数:7
相关论文
共 50 条
  • [1] Towards Understanding and Mitigating Social Biases in Language Models
    Liang, Paul Pu
    Wu, Chiyu
    Morency, Louis-Philippe
    Salakhutdinov, Ruslan
    [J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 139, 2021, 139
  • [2] MODELS OF NATURAL-LANGUAGE UNDERSTANDING
    BATES, M
    [J]. PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 1995, 92 (22) : 9977 - 9982
  • [3] Predicting Garden Path Sentences Based on Natural Language Understanding System
    Du Jia-li
    Yu Ping-fang
    [J]. INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS, 2012, 3 (11) : 1 - 6
  • [4] Understanding Legal Documents: Classification of Rhetorical Role of Sentences Using Deep Learning and Natural Language Processing
    Ahmad, Rameel
    Harris, Deborah
    Sahibzada, Mohammad Ibrahim
    [J]. 2020 IEEE 14TH INTERNATIONAL CONFERENCE ON SEMANTIC COMPUTING (ICSC 2020), 2020, : 464 - 467
  • [5] Fertility models for statistical natural language understanding
    Della Pietra, S
    Epstein, M
    Roukos, S
    Ward, T
    [J]. 35TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 8TH CONFERENCE OF THE EUROPEAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, PROCEEDINGS OF THE CONFERENCE, 1997, : 168 - 173
  • [6] Understanding Natural Language Sentences with Word Embedding and Multi-modal Interaction
    Zhong, Junpei
    Ogata, Tetsuya
    Cangelosi, Angelo
    Yang, Chenguang
    [J]. 2017 THE SEVENTH JOINT IEEE INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING AND EPIGENETIC ROBOTICS (ICDL-EPIROB), 2017, : 184 - 189
  • [7] Canary Extraction in Natural Language Understanding Models
    Parikh, Rahil
    Dupuy, Christophe
    Gupta, Rahul
    [J]. PROCEEDINGS OF THE 60TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2022): (SHORT PAPERS), VOL 2, 2022, : 552 - 560
  • [8] Shortcut Learning of Large Language Models in Natural Language Understanding
    Du, Mengnan
    He, Fengxiang
    Zou, Na
    Tao, Dacheng
    Hu, Xia
    [J]. COMMUNICATIONS OF THE ACM, 2024, 67 (01) : 110 - 120
  • [9] Language understanding using hidden understanding models
    Schwartz, R
    Miller, S
    Stallard, D
    Makhoul, J
    [J]. ICSLP 96 - FOURTH INTERNATIONAL CONFERENCE ON SPOKEN LANGUAGE PROCESSING, PROCEEDINGS, VOLS 1-4, 1996, : 997 - 1000
  • [10] UNDERSTANDING OF SENTENCES IN A RESTRICTED RUSSIAN LANGUAGE
    BAKLANOV, VM
    POPOV, EV
    [J]. ENGINEERING CYBERNETICS, 1978, 16 (04): : 52 - 52