[论文] Your Prompt Should Do More: Effects of Retrieval Instructions in Embed...

论文概要

研究领域: NLP 作者: Amanda Myntti, Jenna Kanerva, Veronika Laippala, Filip Ginter 发布时间: 2026-10-07 arXiv: 2610.10508

中文摘要

带提示的嵌入模型近期受到越来越多的关注，特别是在检索任务中——详细的检索指令作为检索提示的一部分提供。几个新数据集和研究考察了这一设置，表明当前嵌入模型往往难以可靠地遵循此类指令。本文研究指令究竟如何影响非对称检索任务中查询表示的机制。我们表明，当评估中包含查询侧干扰项时，模型甚至无法遵循简单的任务指令。我们假设这种行为源于当前嵌入模型的训练设置和评估方式，并表明使用查询侧干扰项进行微调可带来显著改进，同时对其他任务影响很小。

原文摘要

Prompted embedding models have recently received increasing attention, particularly for retrieval, where detailed retrieval instructions are provided as part of the retrieval prompt. Several new datasets and studies have examined this setting, showing that the current embedding models often struggle to follow such instructions reliably. In this paper, we study the mechanism of how instructions actually affect the representations of retrieval queries in asymmetric retrieval tasks. We show that models can fail to follow even simple task instructions when query-side distractors are included in the evaluation. We hypothesize that this behavior is driven by the training setup of current embedding models and their evaluation, and show that fine-tuning with added query-side distractors leads to s...

#论文 #arXiv #NLP #小凯