{"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/interactive-grounded-language-acquisition-and","title":"Interactive Grounded Language Acquisition and Generalization in a 2D World","arxiv_id":"1802.01433","date":"2018-01-31","proceeding":"ICLR 2018 1","authors":["Haonan Yu","Haichao Zhang","Wei Xu"],"abstract":"We build a virtual agent for learning language in a 2D maze-like world. The\nagent sees images of the surrounding environment, listens to a virtual teacher,\nand takes actions to receive rewards. It interactively learns the teacher's\nlanguage from scratch based on two language use cases: sentence-directed\nnavigation and question answering. It learns simultaneously the visual\nrepresentations of the world, the language, and the action control. By\ndisentangling language grounding from other computational routines and sharing\na concept detection function between language grounding and prediction, the\nagent reliably interpolates and extrapolates to interpret sentences that\ncontain new word combinations or new words missing from training sentences. The\nnew words are transferred from the answers of language prediction. Such a\nlanguage ability is trained and evaluated on a population of over 1.6 million\ndistinct sentences consisting of 119 object words, 8 color words, 9\nspatial-relation words, and 50 grammatical words. The proposed model\nsignificantly outperforms five comparison methods for interpreting zero-shot\nsentences. In addition, we demonstrate human-interpretable intermediate outputs\nof the model in the appendix.","url_abs":"http://arxiv.org/abs/1802.01433v4","url_pdf":"http://arxiv.org/pdf/1802.01433v4.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":"interactive-grounded-language-acquisition-and","repo_url":"https://github.com/PaddlePaddle/XWorld","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"paddle","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"interactive-grounded-language-acquisition-and","repo_url":"https://github.com/ZHEvent/ZHEvent.github.io","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"language-acquisition","task_name":"Language Acquisition"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.01433","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}