arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2408.11865cs.CLcs.AIcs.LG
收藏

LLMs 在提示词中受影响的程度有多大?

How Susceptible are LLMs to Influence in Prompts?

ETH Zürich(苏黎世联邦理工学院);Google DeepMind(谷歌DeepMind)

Sotiris Anagnostidis, Jannis Bulian

AI总结:

本研究探讨了 LLMs 在提示词中受外部模型补充信息影响的程度,发现模型极易受解释左右且不论质量,权威性或高置信度影响较小,强调了审查外部输出的必要性。

摘要:

大型语言模型(LLMs)对提示词(包括其中提供的额外上下文)高度敏感。随着 LLMs 能力的增强,理解其对提示词的敏感性对于确保可靠且稳健的性能变得越来越关键,特别是由于评估这些模型变得更加具有挑战性。在本研究中,我们调查了当前模型(Llama、Mixtral、Falcon)在接收到来自另一个模型的额外输入时的响应,这模拟了一个场景,即一个能力更强的模型——或一个能访问更多外部信息的系统——向目标模型提供补充信息。在多样化的问答任务中,我们研究了当提示词包含来自另一个模型的预测和解释时,LLM 对多项选择题的响应如何变化。具体而言,我们探讨了解释的存在、来源所声称的权威性以及补充输入所声称的置信度的影响。我们的发现表明,模型受到强烈影响,并且当提供解释时,无论解释的质量如何,它们都会被左右。如果输入被呈现为具有权威性或高置信度,模型更容易被左右,但效果在规模上较小。本研究强调了 LLMs 显著的提示词敏感性,并突出了在未经彻底审查和进一步验证的情况下纳入外部来源输出的潜在风险。随着 LLMs 的持续进步,理解和减轻此类敏感性对于其可靠和值得信赖的部署至关重要。

英文摘要:

Large Language Models (LLMs) are highly sensitive to prompts, including additional context provided therein. As LLMs grow in capability, understanding their prompt-sensitivity becomes increasingly crucial for ensuring reliable and robust performance, particularly since evaluating these models becomes more challenging. In this work, we investigate how current models (Llama, Mixtral, Falcon) respond when presented with additional input from another model, mimicking a scenario where a more capable model -- or a system with access to more external information -- provides supplementary information to the target model. Across a diverse spectrum of question-answering tasks, we study how an LLM's response to multiple-choice questions changes when the prompt includes a prediction and explanation from another model. Specifically, we explore the influence of the presence of an explanation, the stated authoritativeness of the source, and the stated confidence of the supplementary input. Our findings reveal that models are strongly influenced, and when explanations are provided they are swayed irrespective of the quality of the explanation. The models are more likely to be swayed if the input is presented as being authoritative or confident, but the effect is small in size. This study underscores the significant prompt-sensitivity of LLMs and highlights the potential risks of incorporating outputs from external sources without thorough scrutiny and further validation. As LLMs continue to advance, understanding and mitigating such sensitivities will be crucial for their reliable and trustworthy deployment.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑