Scholarlink

<p>Using few-shot demonstrations in prompts significantly enhances the generation quality of large language models (LLMs), including code generation. However, adversarial examples injected by malicious service providers via few-shot prompting pose a risk of backdoor attacks in large language models. There is no research on backdoor attacks on large language models in the few-shot prompting setting for code generation tasks. In this paper, we propose <Emphasis FontCategory="NonProportional">BadCodePrompt</Emphasis>, the first backdoor attack for code generation tasks targeting LLMS in the few-shot prompting scenario, without requiring access to training data or model parameters and with lower computational overhead. <Emphasis FontCategory="NonProportional">BadCodePrompt</Emphasis> exploits the insertion of triggers and poisonous code patterns into examples, causing the output of poisonous source code when there is a backdoor trigger in the end user’s query prompt. We demonstrate the effectiveness of <Emphasis FontCategory="NonProportional">BadCodePrompt</Emphasis> in conducting backdoor attacks on three LLMS (GPT-4, Claude-3.5-Sonnet, and Gemini Pro-1.5) in code generation tasks without affecting the functionality of the generated code. LLMs with stronger reasoning capabilities are also more vulnerable to <Emphasis FontCategory="NonProportional">BadCodePrompt</Emphasis>, with an average attack success rate of up to 98.53% for GPT-4 in two benchmark tasks. Finally, we employ state-of-the-art defenses against backdoor attacks in Prompt Engineering and show their overall ineffectiveness against <Emphasis FontCategory="NonProportional">BadCodePrompt</Emphasis>. Therefore, <Emphasis FontCategory="NonProportional">BadCodePrompt</Emphasis> remains a serious threat to LLMS, underscoring the urgency of developing effective defense mechanisms.</p>

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

BadCodePrompt: backdoor attacks against prompt engineering of large language models for code generation

  • 期刊论文

  • 发表日期：

    2025年1月28日

查看全文
  • Yubin Qu,
  • Song Huang,
  • Yanzhou Li,
  • Tongtong Bai,
  • Xiang Chen,
  • Xingya Wang,
  • Long Li,
  • Yongming Yao

摘要

Using few-shot demonstrations in prompts significantly enhances the generation quality of large language models (LLMs), including code generation. However, adversarial examples injected by malicious service providers via few-shot prompting pose a risk of backdoor attacks in large language models. There is no research on backdoor attacks on large language models in the few-shot prompting setting for code generation tasks. In this paper, we propose BadCodePrompt, the first backdoor attack for code generation tasks targeting LLMS in the few-shot prompting scenario, without requiring access to training data or model parameters and with lower computational overhead. BadCodePrompt exploits the insertion of triggers and poisonous code patterns into examples, causing the output of poisonous source code when there is a backdoor trigger in the end user’s query prompt. We demonstrate the effectiveness of BadCodePrompt in conducting backdoor attacks on three LLMS (GPT-4, Claude-3.5-Sonnet, and Gemini Pro-1.5) in code generation tasks without affecting the functionality of the generated code. LLMs with stronger reasoning capabilities are also more vulnerable to BadCodePrompt, with an average attack success rate of up to 98.53% for GPT-4 in two benchmark tasks. Finally, we employ state-of-the-art defenses against backdoor attacks in Prompt Engineering and show their overall ineffectiveness against BadCodePrompt. Therefore, BadCodePrompt remains a serious threat to LLMS, underscoring the urgency of developing effective defense mechanisms.