{"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/explaining-nonlinear-classification-decisions","title":"Explaining NonLinear Classification Decisions with Deep Taylor Decomposition","arxiv_id":"1512.02479","date":"2015-12-08","proceeding":null,"authors":["Grégoire Montavon","Sebastian Bach","Alexander Binder","Wojciech Samek","Klaus-Robert Müller"],"abstract":"Nonlinear methods such as Deep Neural Networks (DNNs) are the gold standard\nfor various challenging machine learning problems, e.g., image classification,\nnatural language processing or human action recognition. Although these methods\nperform impressively well, they have a significant disadvantage, the lack of\ntransparency, limiting the interpretability of the solution and thus the scope\nof application in practice. Especially DNNs act as black boxes due to their\nmultilayer nonlinear structure. In this paper we introduce a novel methodology\nfor interpreting generic multilayer neural networks by decomposing the network\nclassification decision into contributions of its input elements. Although our\nfocus is on image classification, the method is applicable to a broad set of\ninput data, learning tasks and network architectures. Our method is based on\ndeep Taylor decomposition and efficiently utilizes the structure of the network\nby backpropagating the explanations from the output to the input layer. We\nevaluate the proposed method empirically on the MNIST and ILSVRC data sets.","url_abs":"http://arxiv.org/abs/1512.02479v1","url_pdf":"http://arxiv.org/pdf/1512.02479v1.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":"explaining-nonlinear-classification-decisions","repo_url":"https://github.com/OpenXAIProject/LRP_for_LSTM_Korean_dataset","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"explaining-nonlinear-classification-decisions","repo_url":"https://github.com/ginkyenglee/Explaining_Decision_of_Time_Series_Data","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"explaining-nonlinear-classification-decisions","repo_url":"https://github.com/myc159/Deep-Taylor-Decomposition","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"explaining-nonlinear-classification-decisions","repo_url":"https://github.com/taolicheng/understanding-dnn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"action-recognition-in-videos","task_name":"Action Recognition"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"action-recognition","task_name":"Temporal Action Localization"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"interpretability","method_name":"Interpretability"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1512.02479","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1512.02479"}},"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/myc159/Deep-Taylor-Decomposition","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/OpenXAIProject/LRP_for_LSTM_Korean_dataset","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/taolicheng/understanding-dnn","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ginkyenglee/Explaining_Decision_of_Time_Series_Data","reach":{"status":"ok"}}],"summary":{"unverified":1},"by_repo_kind":{"listed":{"samples":1,"ran":0,"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":1,"samples":[{"code_sha256_prefix":"1cfbefabc6650ad2","entry":"model_flattening","repo":"myc159/Deep-Taylor-Decomposition","repo_kind":"listed","path":"model/saliency_mapping.py","file_url":"https://github.com/myc159/Deep-Taylor-Decomposition/blob/HEAD/model/saliency_mapping.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1cfbefabc6650ad2"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}