Harnessing the Universal Geometry of Embeddings
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915754070245376 |
|---|---|
| author | Jha, Rishi Zhang, Collin Shmatikov, Vitaly Morris, John X. |
| author_facet | Jha, Rishi Zhang, Collin Shmatikov, Vitaly Morris, John X. |
| contents | We introduce the first method for translating text embeddings from one vector space to another without any paired data, encoders, or predefined sets of matches. Our unsupervised approach translates any embedding to and from a universal latent representation (i.e., a universal semantic structure conjectured by the Platonic Representation Hypothesis). Our translations achieve high cosine similarity across model pairs with different architectures, parameter counts, and training datasets.
The ability to translate unknown embeddings into a different space while preserving their geometry has serious implications for the security of vector databases. An adversary with access only to embedding vectors can extract sensitive information about the underlying documents, sufficient for classification and attribute inference. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_12540 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Harnessing the Universal Geometry of Embeddings Jha, Rishi Zhang, Collin Shmatikov, Vitaly Morris, John X. Machine Learning We introduce the first method for translating text embeddings from one vector space to another without any paired data, encoders, or predefined sets of matches. Our unsupervised approach translates any embedding to and from a universal latent representation (i.e., a universal semantic structure conjectured by the Platonic Representation Hypothesis). Our translations achieve high cosine similarity across model pairs with different architectures, parameter counts, and training datasets. The ability to translate unknown embeddings into a different space while preserving their geometry has serious implications for the security of vector databases. An adversary with access only to embedding vectors can extract sensitive information about the underlying documents, sufficient for classification and attribute inference. |
| title | Harnessing the Universal Geometry of Embeddings |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2505.12540 |