Scholarlink

Numerous studies have highlighted the privacy risks associated with pretrained large language models. This chapter offers a unique perspective by demonstrating that pretrained large language models can effectively contribute to privacy preservation. We present a locally differentially private mechanism called DP-Prompt proposed in Utpala et al. (Locally differentially private document generation using zero shot prompting. In: Findings of the Association for Computational Linguistics: EMNLP 2023, pp. 8442–8457. Association for Computational Linguistics (2023)), which leverages the power of pretrained large language models and zero-shot prompting to counter author de-anonymization attacks while minimizing the impact on downstream utility.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prompting Large Language Models with Privacy

  • 章节

  • 发表日期：

    2025年

查看全文
  • Pin-Yu Chen,
  • Sijia Liu

摘要

Numerous studies have highlighted the privacy risks associated with pretrained large language models. This chapter offers a unique perspective by demonstrating that pretrained large language models can effectively contribute to privacy preservation. We present a locally differentially private mechanism called DP-Prompt proposed in Utpala et al. (Locally differentially private document generation using zero shot prompting. In: Findings of the Association for Computational Linguistics: EMNLP 2023, pp. 8442–8457. Association for Computational Linguistics (2023)), which leverages the power of pretrained large language models and zero-shot prompting to counter author de-anonymization attacks while minimizing the impact on downstream utility.