{"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/essentially-no-barriers-in-neural-network","title":"Essentially No Barriers in Neural Network Energy Landscape","arxiv_id":"1803.00885","date":"2018-03-02","proceeding":"ICML 2018 7","authors":["Felix Draxler","Kambis Veschgini","Manfred Salmhofer","Fred A. Hamprecht"],"abstract":"Training neural networks involves finding minima of a high-dimensional\nnon-convex loss function. Knowledge of the structure of this energy landscape\nis sparse. Relaxing from linear interpolations, we construct continuous paths\nbetween minima of recent neural network architectures on CIFAR10 and CIFAR100.\nSurprisingly, the paths are essentially flat in both the training and test\nlandscapes. This implies that neural networks have enough capacity for\nstructural changes, or that these changes are small between minima. Also, each\nminimum has at least one vanishing Hessian eigenvalue in addition to those\nresulting from trivial invariance.","url_abs":"http://arxiv.org/abs/1803.00885v5","url_pdf":"http://arxiv.org/pdf/1803.00885v5.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":"essentially-no-barriers-in-neural-network","repo_url":"https://github.com/fdraxler/PyTorch-AutoNEB","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"essentially-no-barriers-in-neural-network","repo_url":"https://github.com/g-benton/loss-surface-simplexes","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.00885","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}