@inproceedings{tehenan-2025-semantic,
    title = "Semantic Geometry of Sentence Embeddings",
    author = "Tehenan, Matthieu",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/name-variant-enfa-fane/2025.findings-emnlp.641/",
    doi = "10.18653/v1/2025.findings-emnlp.641",
    pages = "11993--12004",
    ISBN = "979-8-89176-335-7",
    abstract = "Sentence embeddings are central to modern natural language processing, powering tasks such as clustering, semantic search, and retrieval-augmented generation. Yet, they remain largely opaque: their internal features are not directly interpretable, and users lack fine-grained control for downstream tasks. To address this issue, we introduce a formal framework to characterize the organization of features in sentence embeddings through information-theoretic means. Building on this foundation, we develop a method to identify interpretable feature directions and show how they can be composed to capture richer semantic structures. Experiments on both synthetic and real-world datasets confirm the presence of this semantic geometry and highlight the utility of our approach for enhancing interpretability and fine-grained control in sentence embeddings."
}