{"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/assessing-composition-in-sentence-vector","title":"Assessing Composition in Sentence Vector Representations","arxiv_id":"1809.03992","date":"2018-09-11","proceeding":"COLING 2018 8","authors":["Allyson Ettinger","Ahmed Elgohary","Colin Phillips","Philip Resnik"],"abstract":"An important component of achieving language understanding is mastering the\ncomposition of sentence meaning, but an immediate challenge to solving this\nproblem is the opacity of sentence vector representations produced by current\nneural sentence composition models. We present a method to address this\nchallenge, developing tasks that directly target compositional meaning\ninformation in sentence vector representations with a high degree of precision\nand control. To enable the creation of these controlled tasks, we introduce a\nspecialized sentence generation system that produces large, annotated sentence\nsets meeting specified syntactic, semantic and lexical constraints. We describe\nthe details of the method and generation system, and then present results of\nexperiments applying our method to probe for compositional information in\nembeddings from a number of existing sentence composition models. We find that\nthe method is able to extract useful information about the differing capacities\nof these models, and we discuss the implications of our results with respect to\nthese systems' capturing of sentence information. We make available for public\nuse the datasets used for these experiments, as well as the generation system.","url_abs":"http://arxiv.org/abs/1809.03992v1","url_pdf":"http://arxiv.org/pdf/1809.03992v1.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":"assessing-composition-in-sentence-vector","repo_url":"https://github.com/aetting/compeval-generation-system","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1809.03992","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}