{"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/adversarial-example-generation-with","title":"Adversarial Example Generation with Syntactically Controlled Paraphrase Networks","arxiv_id":"1804.06059","date":"2018-04-17","proceeding":"NAACL 2018 6","authors":["Mohit Iyyer","John Wieting","Kevin Gimpel","Luke Zettlemoyer"],"abstract":"We propose syntactically controlled paraphrase networks (SCPNs) and use them\nto generate adversarial examples. Given a sentence and a target syntactic form\n(e.g., a constituency parse), SCPNs are trained to produce a paraphrase of the\nsentence with the desired syntax. We show it is possible to create training\ndata for this task by first doing backtranslation at a very large scale, and\nthen using a parser to label the syntactic transformations that naturally occur\nduring this process. Such data allows us to train a neural encoder-decoder\nmodel with extra inputs to specify the target syntax. A combination of\nautomated and human evaluations show that SCPNs generate paraphrases that\nfollow their target specifications without decreasing paraphrase quality when\ncompared to baseline (uncontrolled) paraphrase systems. Furthermore, they are\nmore capable of generating syntactically adversarial examples that both (1)\n\"fool\" pretrained models and (2) improve the robustness of these models to\nsyntactic variation when used to augment their training data.","url_abs":"http://arxiv.org/abs/1804.06059v1","url_pdf":"http://arxiv.org/pdf/1804.06059v1.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":"adversarial-example-generation-with","repo_url":"https://github.com/miyyer/scpn","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"adversarial-example-generation-with","repo_url":"https://github.com/tducnguyen-ccg/paraphrasing_SCPN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1804.06059","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}