Robust Prompt Optimization for Large Language Models Against Distribution Shifts

被引:0
|
作者
Li, Moxin [1 ]
Wang, Wenjie [1 ]
Feng, Fuli [2 ,3 ]
Cao, Yixin [4 ]
Zhang, Jizhi [2 ]
Chua, Tat-Seng [1 ]
机构
[1] Natl Univ Singapore, Singapore, Singapore
[2] Univ Sci & Technol China, Hefei, Peoples R China
[3] Inst Dataspace, Hefei, Anhui, Peoples R China
[4] Singapore Management Univ, Singapore, Singapore
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
Large Language Model (LLM) has demonstrated significant ability in various Natural Language Processing tasks. However, their effectiveness is highly dependent on the phrasing of the task prompt, leading to research on automatic prompt optimization using labeled task data. We reveal that these prompt optimization techniques are vulnerable to distribution shifts such as subpopulation shifts, which are common for LLMs in real-world scenarios such as customer reviews analysis. In this light, we propose a new problem of robust prompt optimization for LLMs against distribution shifts, which requires the prompt optimized over the labeled source group can simultaneously generalize to an unlabeled target group. To solve this problem, we propose Generalized Prompt Optimization framework, which incorporates the unlabeled data from the target group into prompt optimization. Extensive experimental results demonstrate the effectiveness of the proposed framework with significant performance improvement on the target group and comparable performance on the source group.
引用
收藏
页码:1539 / 1554
页数:16
相关论文
共 50 条
  • [41] Turning Large Language Models into AI Assistants for Startups Using Prompt Patterns
    Wang, Xiaofeng
    Attal, Mohammad Idris
    Rafiq, Usman
    Hubner-Benz, Sylvia
    AGILE PROCESSES IN SOFTWARE ENGINEERING AND EXTREME PROGRAMMING - WORKSHOPS, XP 2022 WORKSHOPS, XP 2023 WORKSHOPS, 2024, 489 : 192 - 200
  • [42] Bounding the Capabilities of Large Language Models in Open Text Generation with Prompt Constraints
    Lu, Albert
    Zhang, Hongxin
    Zhang, Yanzhe
    Wang, Xuezhi
    Yang, Diyi
    17TH CONFERENCE OF THE EUROPEAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, EACL 2023, 2023, : 1982 - 2008
  • [43] Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigm
    Reynolds, Laria
    McDonell, Kyle
    EXTENDED ABSTRACTS OF THE 2021 CHI CONFERENCE ON HUMAN FACTORS IN COMPUTING SYSTEMS (CHI'21), 2021,
  • [44] A Study on Prompt Types for Harmlessness Assessment of Large-Scale Language Models
    Shin, Yejin
    Kim, Song-yi
    Byun, Eun Young
    HCI INTERNATIONAL 2024 POSTERS, PT VII, HCII 2024, 2024, 2120 : 228 - 233
  • [45] A Transcription Prompt-based Efficient Audio Large Language Model for Robust Speech Recognition
    Li, Yangze
    Wang, Xiong
    Cao, Songjun
    Zhang, Yike
    Ma, Long
    Xie, Lei
    INTERSPEECH 2024, 2024, : 1905 - 1909
  • [46] ROBBIE: Robust Bias Evaluation of Large Generative Language Models
    Esiobu, David
    Tan, Xiaoqing
    Hosseini, Saghar
    Ung, Megan
    Zhang, Yuchen
    Fernandes, Jude
    Dwivedi-Yu, Jane
    Presani, Eleonora
    Williams, Adina
    Meta, Eric Michael Smith
    2023 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING, EMNLP 2023, 2023, : 3764 - 3814
  • [47] MAPO: Boosting Large Language Model Performance with Model-Adaptive Prompt Optimization
    Chen, Yuyan
    Wen, Zhihao
    Fan, Ge
    Chen, Zhengyu
    Wu, Wei
    Liu, Dayiheng
    Li, Zhixu
    Liu, Bang
    Xiao, Yanghua
    FINDINGS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS - EMNLP 2023, 2023, : 3279 - 3304
  • [48] Co-training Improves Prompt-based Learning for Large Language Models
    Lang, Hunter
    Agrawal, Monica
    Kim, Yoon
    Sontag, David
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 162, 2022,
  • [49] Large Language Models and Applications: The Rebirth of Enterprise Knowledge Management and the Rise of Prompt Libraries
    O'Leary, Daniel E.
    IEEE INTELLIGENT SYSTEMS, 2024, 39 (02) : 72 - 75
  • [50] A prompt construction method for the reverse dictionary task of large-scale language models
    Tian, Sicheng
    Huang, Shaobin
    Li, Rongsheng
    Wei, Chi
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2024, 133