{"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/tensor-product-generation-networks-for-deep","title":"Tensor Product Generation Networks for Deep NLP Modeling","arxiv_id":"1709.09118","date":"2017-09-26","proceeding":"NAACL 2018 6","authors":["Qiuyuan Huang","Paul Smolensky","Xiaodong He","Li Deng","Dapeng Wu"],"abstract":"We present a new approach to the design of deep networks for natural language\nprocessing (NLP), based on the general technique of Tensor Product\nRepresentations (TPRs) for encoding and processing symbol structures in\ndistributed neural networks. A network architecture --- the Tensor Product\nGeneration Network (TPGN) --- is proposed which is capable in principle of\ncarrying out TPR computation, but which uses unconstrained deep learning to\ndesign its internal representations. Instantiated in a model for image-caption\ngeneration, TPGN outperforms LSTM baselines when evaluated on the COCO dataset.\nThe TPR-capable structure enables interpretation of internal representations\nand operations, which prove to contain considerable grammatical content. Our\ncaption-generation model can be interpreted as generating sequences of\ngrammatical categories and retrieving words by their categories from a plan\nencoded as a distributed representation.","url_abs":"http://arxiv.org/abs/1709.09118v5","url_pdf":"http://arxiv.org/pdf/1709.09118v5.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":"tensor-product-generation-networks-for-deep","repo_url":"https://github.com/ggeorgea/TPRcaption","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"tensor-product-generation-networks-for-deep","repo_url":"https://github.com/tingshengtan/DeepLearningForImageCaptioning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"caption-generation","task_name":"Caption Generation"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}