{"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/towards-deep-neural-network-architectures","title":"Towards Deep Neural Network Architectures Robust to Adversarial Examples","arxiv_id":"1412.5068","date":"2014-12-11","proceeding":null,"authors":["Shixiang Gu","Luca Rigazio"],"abstract":"Recent work has shown deep neural networks (DNNs) to be highly susceptible to\nwell-designed, small perturbations at the input layer, or so-called adversarial\nexamples. Taking images as an example, such distortions are often\nimperceptible, but can result in 100% mis-classification for a state of the art\nDNN. We study the structure of adversarial examples and explore network\ntopology, pre-processing and training strategies to improve the robustness of\nDNNs. We perform various experiments to assess the removability of adversarial\nexamples by corrupting with additional noise and pre-processing with denoising\nautoencoders (DAEs). We find that DAEs can remove substantial amounts of the\nadversarial noise. How- ever, when stacking the DAE with the original DNN, the\nresulting network can again be attacked by new adversarial examples with even\nsmaller distortion. As a solution, we propose Deep Contractive Network, a model\nwith a new end-to-end training procedure that includes a smoothness penalty\ninspired by the contractive autoencoder (CAE). This increases the network\nrobustness to adversarial examples, without a significant performance penalty.","url_abs":"http://arxiv.org/abs/1412.5068v4","url_pdf":"http://arxiv.org/pdf/1412.5068v4.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":"towards-deep-neural-network-architectures","repo_url":"https://github.com/aaronmckinstry706/toward-robust-dnn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"towards-deep-neural-network-architectures","repo_url":"https://github.com/ypotdevin/randomized-defenses","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"}],"methods":[{"method_slug":"contractive-autoencoder","method_name":"Contractive Autoencoder"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1412.5068","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1412.5068"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ypotdevin/randomized-defenses","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/aaronmckinstry706/toward-robust-dnn","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":1,"unverified":1},"by_repo_kind":{"listed":{"samples":2,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"8d0f77459efa8fd3","entry":"aggregate_predict_n_by","repo":"ypotdevin/randomized-defenses","repo_kind":"listed","path":"defense_mechanisms/base.py","file_url":"https://github.com/ypotdevin/randomized-defenses/blob/HEAD/defense_mechanisms/base.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8d0f77459efa8fd3"}},{"code_sha256_prefix":"1c87d38484ebf52b","entry":"curried_aggregate_predict_n_by","repo":"ypotdevin/randomized-defenses","repo_kind":"listed","path":"defense_mechanisms/base.py","file_url":"https://github.com/ypotdevin/randomized-defenses/blob/HEAD/defense_mechanisms/base.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1c87d38484ebf52b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}