{"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/tunneling-neural-perception-and-logic","title":"Tunneling Neural Perception and Logic Reasoning through Abductive Learning","arxiv_id":"1802.01173","date":"2018-02-04","proceeding":null,"authors":["Wang-Zhou Dai","Qiu-Ling Xu","Yang Yu","Zhi-Hua Zhou"],"abstract":"Perception and reasoning are basic human abilities that are seamlessly\nconnected as part of human intelligence. However, in current machine learning\nsystems, the perception and reasoning modules are incompatible. Tasks requiring\njoint perception and reasoning ability are difficult to accomplish autonomously\nand still demand human intervention. Inspired by the way language experts\ndecoded Mayan scripts by joining two abilities in an abductive manner, this\npaper proposes the abductive learning framework. The framework learns\nperception and reasoning simultaneously with the help of a trial-and-error\nabductive process. We present the Neural-Logical Machine as an implementation\nof this novel learning framework. We demonstrate that--using human-like\nabductive learning--the machine learns from a small set of simple hand-written\nequations and then generalizes well to complex equations, a feat that is beyond\nthe capability of state-of-the-art neural network models. The abductive\nlearning framework explores a new direction for approaching human-level\nlearning ability.","url_abs":"http://arxiv.org/abs/1802.01173v2","url_pdf":"http://arxiv.org/pdf/1802.01173v2.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":"tunneling-neural-perception-and-logic","repo_url":"https://github.com/AbductiveLearning/ABL-HED","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.01173","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}