{"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/freezing-subnetworks-to-analyze-domain","title":"Freezing Subnetworks to Analyze Domain Adaptation in Neural Machine Translation","arxiv_id":"1809.05218","date":"2018-09-14","proceeding":"WS 2018 10","authors":["Brian Thompson","Huda Khayrallah","Antonios Anastasopoulos","Arya D. McCarthy","Kevin Duh","Rebecca Marvin","Paul McNamee","Jeremy Gwinnup","Tim Anderson","Philipp Koehn"],"abstract":"To better understand the effectiveness of continued training, we analyze the\nmajor components of a neural machine translation system (the encoder, decoder,\nand each embedding space) and consider each component's contribution to, and\ncapacity for, domain adaptation. We find that freezing any single component\nduring continued training has minimal impact on performance, and that\nperformance is surprisingly good when a single component is adapted while\nholding the rest of the model fixed. We also find that continued training does\nnot move the model very far from the out-of-domain model, compared to a\nsensitivity analysis metric, suggesting that the out-of-domain model can\nprovide a good generic initialization for the new domain.","url_abs":"http://arxiv.org/abs/1809.05218v4","url_pdf":"http://arxiv.org/pdf/1809.05218v4.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":"freezing-subnetworks-to-analyze-domain","repo_url":"https://github.com/awslabs/sockeye","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"mxnet","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1809.05218","atlas_url":"https://app.syntology.ai/?focus=1809.05218","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}