{"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/on-visual-hallmarks-of-robustness-to","title":"On Visual Hallmarks of Robustness to Adversarial Malware","arxiv_id":"1805.03553","date":"2018-05-09","proceeding":null,"authors":["Alex Huang","Abdullah Al-Dujaili","Erik Hemberg","Una-May O'Reilly"],"abstract":"A central challenge of adversarial learning is to interpret the resulting\nhardened model. In this contribution, we ask how robust generalization can be\nvisually discerned and whether a concise view of the interactions between a\nhardened decision map and input samples is possible. We first provide a means\nof visually comparing a hardened model's loss behavior with respect to the\nadversarial variants generated during training versus loss behavior with\nrespect to adversarial variants generated from other sources. This allows us to\nconfirm that the association of observed flatness of a loss landscape with\ngeneralization that is seen with naturally trained models extends to\nadversarially hardened models and robust generalization. To complement these\nmeans of interpreting model parameter robustness we also use self-organizing\nmaps to provide a visual means of superimposing adversarial and natural\nvariants on a model's decision space, thus allowing the model's global\nrobustness to be comprehensively examined.","url_abs":"http://arxiv.org/abs/1805.03553v1","url_pdf":"http://arxiv.org/pdf/1805.03553v1.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":"on-visual-hallmarks-of-robustness-to","repo_url":"https://github.com/ALFA-group/robust-adv-malware-detection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}