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
  • [21] Select, Prompt, Filter: Distilling Large Language Models for Summarizing Conversations
    Pham, Minh-Quang
    Indurthi, Sathish Reddy
    Chollampatt, Shamil
    Turchi, Marco
    2023 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2023), 2023, : 12257 - 12265
  • [22] Integrating chemistry knowledge in large language models via prompt engineering
    Liu, Hongxuan
    Yin, Haoyu
    Luo, Zhiyao
    Wang, Xiaonan
    SYNTHETIC AND SYSTEMS BIOTECHNOLOGY, 2025, 10 (01) : 23 - 38
  • [23] Assessing the Impact of Prompt Strategies on Text Summarization with Large Language Models
    Onan, Aytug
    Alhumyani, Hesham
    COMPUTER APPLICATIONS IN INDUSTRY AND ENGINEERING, CAINE 2024, 2025, 2242 : 41 - 55
  • [24] Soft prompt tuning for augmenting dense retrieval with large language models
    Peng, Zhiyuan
    Wu, Xuyang
    Wang, Qifan
    Fang, Yi
    KNOWLEDGE-BASED SYSTEMS, 2025, 309
  • [25] Prompt Wrangling: On Replication and Generalization in Large Language Models for PCG Levels
    Karkaj, Arash Moradi
    Nelson, Mark J.
    Koutis, Ioannis
    Hoover, Amy K.
    PROCEEDINGS OF THE 19TH INTERNATIONAL CONFERENCE ON THE FOUNDATIONS OF DIGITAL GAMES, FDG 2024, 2024,
  • [26] CSPO: chain-structured prompt optimisation for large language models
    Wang, Jinshui
    Lin, Sining
    Xue, Xingsi
    Chen, Shuguang
    Tang, Zhengyi
    International Journal of Ad Hoc and Ubiquitous Computing, 2025, 48 (04) : 233 - 243
  • [27] Distribution-Aware Prompt Tuning for Vision-Language Models
    Cho, Eulrang
    Kim, Jooyeon
    Kim, Hyunwoo J.
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2023), 2023, : 21947 - 21956
  • [28] You Only Prompt Once: On the Capabilities of Prompt Learning on Large Language Models to Tackle Toxic Content
    He, Xinlei
    Zannettou, Savvas
    Shen, Yun
    Zhang, Yang
    45TH IEEE SYMPOSIUM ON SECURITY AND PRIVACY, SP 2024, 2024, : 770 - 787
  • [29] Exploring Distributional Shifts in Large Language Models for Code Analysis
    Arakelyan, Shushan
    Das, Rocktim Jyoti
    Mao, Yi
    Ren, Xiang
    2023 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2023), 2023, : 16298 - 16314
  • [30] TPIA: Towards Target-specific Prompt Injection Attack against Code-oriented Large Language Models
    Yang, Yuchen
    Yao, Hongwei
    Yang, Bingrun
    He, Yiling
    Li, Yiming
    Zhang, Tianwei
    Qin, Zhan
    Ren, Kui
    Chen, Chun
    arXiv,