{"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/hierarchical-quantized-representations-for","title":"Hierarchical Quantized Representations for Script Generation","arxiv_id":"1808.09542","date":"2018-08-28","proceeding":"EMNLP 2018 10","authors":["Noah Weber","Leena Shekhar","Niranjan Balasubramanian","Nathanael Chambers"],"abstract":"Scripts define knowledge about how everyday scenarios (such as going to a\nrestaurant) are expected to unfold. One of the challenges to learning scripts\nis the hierarchical nature of the knowledge. For example, a suspect arrested\nmight plead innocent or guilty, and a very different track of events is then\nexpected to happen. To capture this type of information, we propose an\nautoencoder model with a latent space defined by a hierarchy of categorical\nvariables. We utilize a recently proposed vector quantization based approach,\nwhich allows continuous embeddings to be associated with each latent variable\nvalue. This permits the decoder to softly decide what portions of the latent\nhierarchy to condition on by attending over the value embeddings for a given\nsetting. Our model effectively encodes and generates scripts, outperforming a\nrecent language modeling-based method on several standard tasks, and allowing\nthe autoencoder model to achieve substantially lower perplexity scores compared\nto the previous language modeling-based method.","url_abs":"http://arxiv.org/abs/1808.09542v1","url_pdf":"http://arxiv.org/pdf/1808.09542v1.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":"hierarchical-quantized-representations-for","repo_url":"https://github.com/StonyBrookNLP/HAQAE","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"quantization","task_name":"Quantization"},{"task_slug":"script-generation","task_name":"Script Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.09542","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}