大语言模型是上下文中的个性化摘要器吗?用 iCOPERNICUS 测试检验一下!
Are Large Language Models In-Context Personalized Summarizers? Get an iCOPERNICUS Test Done!
AI总结:
本文提出 iCOPERNICUS 框架,以 EGISES 为度量检验 LLMs 是否真正具备上下文个性化学习能力,发现 15/17 个模型在更丰富提示下 ICPL 下降,缺乏真正的个性化摘要能力。
摘要:
大语言模型(LLMs)在基于上下文学习(In-Context-Learning,ICL)的摘要任务中已取得显著成功。然而,显著性取决于用户特定的偏好历史。因此,我们需要在此类 LLMs 中具备可靠的上下文个性化学习(In-Context Personalization Learning,ICPL)能力。任意 LLM 要展现 ICPL,都需要具备辨别用户画像中对比差异的能力。最近的一项研究首次提出了一种名为 EGISES 的个性化程度度量方法。EGISES 衡量模型对用户画像差异的响应程度。然而,它无法检验模型是否利用了 ICPL 提示中提供的全部三类线索:(i)示例摘要,(ii)用户阅读历史,以及(iii)用户画像中的对比差异。为解决这一问题,我们提出了 iCOPERNICUS 框架,这是一种新颖的针对 LLMs 摘要能力的上下文个性化学习审视(In-COntext PERsonalization learNIng sCrUtiny)框架,该框架使用 EGISES 作为比较度量。作为案例研究,我们根据所报告的 ICL 性能评估了 17 个最先进的 LLMs,并观察到当使用更丰富的提示进行探测时,有 15 个模型的 ICPL 性能下降(最低:1.6%;最高:3.6%),从而表明它们缺乏真正的 ICPL。
英文摘要:
Large Language Models (LLMs) have succeeded considerably in In-Context-Learning (ICL) based summarization. However, saliency is subject to the users' specific preference histories. Hence, we need reliable In-Context Personalization Learning (ICPL) capabilities within such LLMs. For any arbitrary LLM to exhibit ICPL, it needs to have the ability to discern contrast in user profiles. A recent study proposed a measure for degree-of-personalization called EGISES for the first time. EGISES measures a model's responsiveness to user profile differences. However, it cannot test if a model utilizes all three types of cues provided in ICPL prompts: (i) example summaries, (ii) user's reading histories, and (iii) contrast in user profiles. To address this, we propose the iCOPERNICUS framework, a novel In-COntext PERsonalization learNIng sCrUtiny of Summarization capability in LLMs that uses EGISES as a comparative measure. As a case-study, we evaluate 17 state-of-the-art LLMs based on their reported ICL performances and observe that 15 models' ICPL degrades (min: 1.6%; max: 3.6%) when probed with richer prompts, thereby showing lack of true ICPL.
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。