{"slug": "harnessing-the-universal-geometry-of-embeddings", "title": "Harnessing the Universal Geometry of Embeddings", "summary": "Researchers introduced the first method for translating text embeddings between vector spaces without paired data, encoders, or predefined matches, achieving high cosine similarity across model pairs with different architectures, parameter counts, and training datasets. The method, described in a paper submitted on 18 May 2025 and revised on 26 Jan 2026, raises security concerns for vector databases, as an adversary with access only to embedding vectors could extract sensitive information about underlying documents.", "body_md": "# Computer Science > Machine Learning\n\n  [Submitted on 18 May 2025 (\n\n[v1](https://arxiv.org/abs/2505.12540v1)), last revised 26 Jan 2026 (this version, v4)]\n# Title:Harnessing the Universal Geometry of Embeddings\n\n[View PDF](/pdf/2505.12540)\n\n[HTML (experimental)](https://arxiv.org/html/2505.12540v4)\n\nAbstract: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.\n\nThe 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.\n\n## Submission history\n\nFrom: Rishi Jha [\n[view email](/show-email/465a8825/2505.12540)]\n\n**Sun, 18 May 2025 20:37:07 UTC (3,179 KB)**\n\n[\\[v1\\]](/abs/2505.12540v1)\n**Tue, 20 May 2025 15:38:41 UTC (3,180 KB)**\n\n[\\[v2\\]](/abs/2505.12540v2)\n**Wed, 25 Jun 2025 21:04:02 UTC (2,407 KB)**\n\n[\\[v3\\]](/abs/2505.12540v3)\n**[v4]** Mon, 26 Jan 2026 14:47:13 UTC (2,424 KB)\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\nIArxiv Recommender\n\n*(*[What is IArxiv?](https://iarxiv.org/about))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/harnessing-the-universal-geometry-of-embeddings", "canonical_source": "https://arxiv.org/abs/2505.12540", "published_at": "2026-09-06 20:31:20+00:00", "updated_at": "2026-09-07 02:11:57.265563+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "ai-safety"], "entities": ["arXiv", "Platonic Representation Hypothesis", "Rishi Jha"], "alternates": {"html": "https://wpnews.pro/news/harnessing-the-universal-geometry-of-embeddings", "markdown": "https://wpnews.pro/news/harnessing-the-universal-geometry-of-embeddings.md", "text": "https://wpnews.pro/news/harnessing-the-universal-geometry-of-embeddings.txt", "jsonld": "https://wpnews.pro/news/harnessing-the-universal-geometry-of-embeddings.jsonld"}}