{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/poincare-glove-hyperbolic-word-embeddings","title":"Poincaré GloVe: Hyperbolic Word Embeddings","arxiv_id":"1810.06546","date":"2018-10-15","proceeding":null,"authors":["Alexandru Tifrea","Gary Bécigneul","Octavian-Eugen Ganea"],"abstract":"Words are not created equal. In fact, they form an aristocratic graph with a\nlatent hierarchical structure that the next generation of unsupervised learned\nword embeddings should reveal. In this paper, justified by the notion of\ndelta-hyperbolicity or tree-likeliness of a space, we propose to embed words in\na Cartesian product of hyperbolic spaces which we theoretically connect to the\nGaussian word embeddings and their Fisher geometry. This connection allows us\nto introduce a novel principled hypernymy score for word embeddings. Moreover,\nwe adapt the well-known Glove algorithm to learn unsupervised word embeddings\nin this type of Riemannian manifolds. We further explain how to solve the\nanalogy task using the Riemannian parallel transport that generalizes vector\narithmetics to this new type of geometry. Empirically, based on extensive\nexperiments, we prove that our embeddings, trained unsupervised, are the first\nto simultaneously outperform strong and popular baselines on the tasks of\nsimilarity, analogy and hypernymy detection. In particular, for word hypernymy,\nwe obtain new state-of-the-art on fully unsupervised WBLESS classification\naccuracy.","url_abs":"http://arxiv.org/abs/1810.06546v2","url_pdf":"http://arxiv.org/pdf/1810.06546v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"poincare-glove-hyperbolic-word-embeddings","repo_url":"https://github.com/alex-tifrea/poincare_glove","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"LGPL-2.1"}}],"tasks":[{"task_slug":"word-embeddings","task_name":"Word Embeddings"}],"methods":[{"method_slug":"glove","method_name":"GloVe"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1810.06546","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}