{"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/neural-network-quine","title":"Neural Network Quine","arxiv_id":"1803.05859","date":"2018-03-15","proceeding":null,"authors":["Oscar Chang","Hod Lipson"],"abstract":"Self-replication is a key aspect of biological life that has been largely\noverlooked in Artificial Intelligence systems. Here we describe how to build\nand train self-replicating neural networks. The network replicates itself by\nlearning to output its own weights. The network is designed using a loss\nfunction that can be optimized with either gradient-based or non-gradient-based\nmethods. We also describe a method we call regeneration to train the network\nwithout explicit optimization, by injecting the network with predictions of its\nown parameters. The best solution for a self-replicating network was found by\nalternating between regeneration and optimization steps. Finally, we describe a\ndesign for a self-replicating neural network that can solve an auxiliary task\nsuch as MNIST image classification. We observe that there is a trade-off\nbetween the network's ability to classify images and its ability to replicate,\nbut training is biased towards increasing its specialization at image\nclassification at the expense of replication. This is analogous to the\ntrade-off between reproduction and other tasks observed in nature. We suggest\nthat a self-replication mechanism for artificial intelligence is useful because\nit introduces the possibility of continual improvement through natural\nselection.","url_abs":"http://arxiv.org/abs/1803.05859v4","url_pdf":"http://arxiv.org/pdf/1803.05859v4.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":"neural-network-quine","repo_url":"https://github.com/AustinT/nn-quine","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":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}