[论文] Embedding Models Measure in Peculiar Ways

论文概要

研究领域: NLP 作者: Juri Opitz, Andrianos Michail 发布时间: 2026-09-17 arXiv: 2609.20821

中文摘要

嵌入空间定义了语义相似性和距离的概念。我们研究这些嵌入是否反映了质量、距离、时间和体积的物理量度——这些量度具有唯一的、客观的语义等价和距离概念。我们发现，物理量度在嵌入空间中仅被微弱建模，反而可以观察到相当奇特的度量模式。进一步分析表明，嵌入对物理量度的表示深受表层字符串相似性的影响，而对相似性进行重新校准并不能实质改善对齐效果。

原文摘要

Embedding spaces define notions of semantic similarity and distance. We study whether those embeddings reflect physical measurements of mass, distance, time and volume, which admit a unique, objective notion of semantic equivalence and distance. We find that physical measurement is only weakly modeled in the embedding space, and that instead quite peculiar measurement patterns can be observed. Further analysis indicates that embedding representations of physical measurements are strongly influenced by superficial string similarity, and recalibration of similarity does not substantially improve the alignment.

#论文 #arXiv #NLP #小凯