{"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-representations-for-efficient","title":"Hierarchical Representations for Efficient Architecture Search","arxiv_id":"1711.00436","date":"2017-11-01","proceeding":"ICLR 2018 1","authors":["Hanxiao Liu","Karen Simonyan","Oriol Vinyals","Chrisantha Fernando","Koray Kavukcuoglu"],"abstract":"We explore efficient neural architecture search methods and show that a\nsimple yet powerful evolutionary algorithm can discover new architectures with\nexcellent performance. Our approach combines a novel hierarchical genetic\nrepresentation scheme that imitates the modularized design pattern commonly\nadopted by human experts, and an expressive search space that supports complex\ntopologies. Our algorithm efficiently discovers architectures that outperform a\nlarge number of manually designed models for image classification, obtaining\ntop-1 error of 3.6% on CIFAR-10 and 20.3% when transferred to ImageNet, which\nis competitive with the best existing neural architecture search approaches. We\nalso present results using random search, achieving 0.3% less top-1 accuracy on\nCIFAR-10 and 0.1% less on ImageNet whilst reducing the search time from 36\nhours down to 1 hour.","url_abs":"http://arxiv.org/abs/1711.00436v2","url_pdf":"http://arxiv.org/pdf/1711.00436v2.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-representations-for-efficient","repo_url":"https://github.com/sjoshi804/hierarchical-darts-neural-architecture-search","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"architecture-search","task_name":"Neural Architecture Search"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.00436","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}