arXiv cs.CL· Kohei Oda, Po-Min Chuang, Kiyoaki Shirai, Natthawut Kertkeidkachorn·· 12 小时前AI 评分20
基于层级问题构建可解释文本嵌入的方法
Interpretable Text Embeddings Based on Hierarchical Questions
AI 导读
研究者提出用层级化问题构建基于 QA 的文本嵌入：随着层级加深，问题更具体、数量更多，解读时用浅层对齐问题替换深层问题以减少人工检查量。该方法通过递归聚类、对比提示词和问题对齐构建问题层级，在通用任务上提升了可解释性并保持任务性能。实验还发现，问题数量有限时浅层问题比深层问题任务表现更好。
正文 · 原文
Abstract:QA-based text embeddings represent texts as vectors consisting of 1/0 answers to yes/no questions. Although these embeddings provide detailed information for interpretation and achieve high task performance, they often require humans to examine a large number of questions. Furthermore, the relationship between the number of questions and the appropriate level of question granularity has not been sufficiently studied. To address these challenges, we propose a method for constructing hierarchical questions for text embeddings. As the hierarchy deepens, questions become more specific and their number increases. During interpretation, we reduce the number of questions that humans need to examine by replacing questions at deeper levels with aligned questions at shallower levels. The question hierarchy is constructed through recursive clustering, contrastive prompting, and question alignment. Experimental results on general-purpose tasks show that our method improves interpretability while maintaining task performance. It is also found that, when only a limited number of questions are available, questions at shallower levels tend to achieve higher task performance than those at deeper levels, even with the same number of questions.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.06402 [cs.CL] |
| (or arXiv:2611.06402v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.06402 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Kohei Oda [view email]
[v1]
Wed, 4 Nov 2026 12:05:15 UTC (3,456 KB)
来源：arXiv cs.CL · arxiv.org