{"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/interpretable-textual-neuron-representations","title":"Interpretable Textual Neuron Representations for NLP","arxiv_id":"1809.07291","date":"2018-09-19","proceeding":"WS 2018 11","authors":["Nina Poerner","Benjamin Roth","Hinrich Schütze"],"abstract":"Input optimization methods, such as Google Deep Dream, create interpretable\nrepresentations of neurons for computer vision DNNs. We propose and evaluate\nways of transferring this technology to NLP. Our results suggest that gradient\nascent with a gumbel softmax layer produces n-gram representations that\noutperform naive corpus search in terms of target neuron activation. The\nrepresentations highlight differences in syntax awareness between the language\nand visual models of the Imaginet architecture.","url_abs":"http://arxiv.org/abs/1809.07291v1","url_pdf":"http://arxiv.org/pdf/1809.07291v1.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":"interpretable-textual-neuron-representations","repo_url":"https://github.com/NPoe/input-optimization-nlp","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"interpretable-textual-neuron-representations","repo_url":"https://github.com/phnk/D7047E","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[],"methods":[{"method_slug":"gumbel-softmax","method_name":"Gumbel Softmax"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1809.07291","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}