{"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/all-you-need-is-a-good-init","title":"All you need is a good init","arxiv_id":"1511.06422","date":"2015-11-19","proceeding":"ICLR 2015 11","authors":["Dmytro Mishkin","Jiri Matas"],"abstract":"Layer-sequential unit-variance (LSUV) initialization - a simple method for\nweight initialization for deep net learning - is proposed. The method consists\nof the two steps. First, pre-initialize weights of each convolution or\ninner-product layer with orthonormal matrices. Second, proceed from the first\nto the final layer, normalizing the variance of the output of each layer to be\nequal to one.\n  Experiment with different activation functions (maxout, ReLU-family, tanh)\nshow that the proposed initialization leads to learning of very deep nets that\n(i) produces networks with test accuracy better or equal to standard methods\nand (ii) is at least as fast as the complex schemes proposed specifically for\nvery deep nets such as FitNets (Romero et al. (2015)) and Highway (Srivastava\net al. (2015)).\n  Performance is evaluated on GoogLeNet, CaffeNet, FitNets and Residual nets\nand the state-of-the-art, or very close to it, is achieved on the MNIST,\nCIFAR-10/100 and ImageNet datasets.","url_abs":"http://arxiv.org/abs/1511.06422v7","url_pdf":"http://arxiv.org/pdf/1511.06422v7.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":"all-you-need-is-a-good-init","repo_url":"https://github.com/ducha-aiki/LSUVinit","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/JonasWechsler/DeepLearningLab5","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/dmbernaal/Daedalus","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/ducha-aiki/LSUV-keras","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/ducha-aiki/LSUV-pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/ducha-aiki/lsuv","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/shunk031/LSUV.pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/vimarshc/fastai_experiments","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/MindCode-4/code-12/tree/main/lsuv-init","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/MindCode-4/code-7/tree/main/lsuv-init","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"all-you-need-is-a-good-init","repo_url":"https://github.com/MindSpore-scientific/code-6/tree/main/lsuv-init","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"all","task_name":"All"},{"task_slug":"image-classification","task_name":"Image Classification"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"auxiliary-classifier","method_name":"Auxiliary Classifier"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"googlenet","method_name":"GoogLeNet"},{"method_slug":"inception-module","method_name":"Inception Module"},{"method_slug":"lsuv-initialization","method_name":"LSUV Initialization"},{"method_slug":"local-response-normalization","method_name":"Local Response Normalization"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-cifar-10","task":"Image Classification","dataset":"CIFAR-10","model":"Fitnet4-LSUV","rank_in_archive_order":161,"of":265,"metrics":{"Percentage correct":"94.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-cifar-100","task":"Image Classification","dataset":"CIFAR-100","model":"Fitnet4-LSUV","rank_in_archive_order":169,"of":211,"metrics":{"Percentage correct":"72.3"},"uses_additional_data":true},{"leaderboard":"/sota/image-classification-on-mnist","task":"Image Classification","dataset":"MNIST","model":"Fitnet-LSUV-SVM","rank_in_archive_order":30,"of":81,"metrics":{"Percentage error":"0.4"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1511.06422","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1511.06422"}},"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/dmbernaal/Daedalus","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/vimarshc/fastai_experiments","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindCode-4/code-7/tree/main/lsuv-init","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ducha-aiki/LSUVinit","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindCode-4/code-12/tree/main/lsuv-init","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/shunk031/LSUV.pytorch","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MindSpore-scientific/code-6/tree/main/lsuv-init","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ducha-aiki/LSUV-keras","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ducha-aiki/lsuv","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ducha-aiki/LSUV-pytorch","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/JonasWechsler/DeepLearningLab5","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":19},"by_repo_kind":{"official":{"samples":1,"ran":0,"repositories":1},"listed":{"samples":18,"ran":0,"repositories":5}},"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":2,"samples":[{"code_sha256_prefix":"a42f70aba8b0e71a","entry":"LSUVinit","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"lsuv_init.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/lsuv_init.py","link_basis":"harvester_set","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":"a42f70aba8b0e71a"}},{"code_sha256_prefix":"522aea77ce74065c","entry":"LSUVinit","repo":"ducha-aiki/LSUV-keras","repo_kind":"listed","path":"lsuv_init.py","file_url":"https://github.com/ducha-aiki/LSUV-keras/blob/HEAD/lsuv_init.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"522aea77ce74065c"}},{"code_sha256_prefix":"02e5a31d6c7a6b1c","entry":"LSUVinit","repo":"ducha-aiki/LSUV-pytorch","repo_kind":"listed","path":"LSUV.py","file_url":"https://github.com/ducha-aiki/LSUV-pytorch/blob/HEAD/LSUV.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"02e5a31d6c7a6b1c"}},{"code_sha256_prefix":"25474dc6ec455a0d","entry":"average_gradients","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"cifar10/cifar10_multi_gpu_train.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/cifar10/cifar10_multi_gpu_train.py","link_basis":"harvester_set","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":"25474dc6ec455a0d"}},{"code_sha256_prefix":"8919d7b64b3fd80a","entry":"distorted_inputs","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"cifar10/cifar10_input.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/cifar10/cifar10_input.py","link_basis":"harvester_set","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":"8919d7b64b3fd80a"}},{"code_sha256_prefix":"f0ac7f8a003439d1","entry":"get_activations","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"lsuv_init.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/lsuv_init.py","link_basis":"harvester_set","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":"f0ac7f8a003439d1"}},{"code_sha256_prefix":"e4da73f8a72851bb","entry":"get_scalable_weight","repo":"ducha-aiki/lsuv","repo_kind":"listed","path":"lsuv/lsuvinit.py","file_url":"https://github.com/ducha-aiki/lsuv/blob/HEAD/lsuv/lsuvinit.py","link_basis":"harvester_set","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":"e4da73f8a72851bb"}},{"code_sha256_prefix":"0f84f9e9bc13c884","entry":"inputs","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"cifar10/cifar10_input.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/cifar10/cifar10_input.py","link_basis":"harvester_set","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":"0f84f9e9bc13c884"}},{"code_sha256_prefix":"6b3581c398b49245","entry":"is_relevant_layer","repo":"ducha-aiki/lsuv","repo_kind":"listed","path":"lsuv/lsuvinit.py","file_url":"https://github.com/ducha-aiki/lsuv/blob/HEAD/lsuv/lsuvinit.py","link_basis":"harvester_set","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":"6b3581c398b49245"}},{"code_sha256_prefix":"7b7d1b28f254c06c","entry":"loss","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"cifar10/cifar10.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/cifar10/cifar10.py","link_basis":"harvester_set","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":"7b7d1b28f254c06c"}},{"code_sha256_prefix":"9c5badcb7115095b","entry":"lsuv_init","repo":"shunk031/LSUV.pytorch","repo_kind":"listed","path":"lsuv/lsuv.py","file_url":"https://github.com/shunk031/LSUV.pytorch/blob/HEAD/lsuv/lsuv.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":"9c5badcb7115095b"}},{"code_sha256_prefix":"1de0deb674c678ec","entry":"move_to","repo":"ducha-aiki/lsuv","repo_kind":"listed","path":"lsuv/lsuvinit.py","file_url":"https://github.com/ducha-aiki/lsuv/blob/HEAD/lsuv/lsuvinit.py","link_basis":"harvester_set","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":"1de0deb674c678ec"}},{"code_sha256_prefix":"c03141891ec617ab","entry":"random_rotation","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"image.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/image.py","link_basis":"harvester_set","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":"c03141891ec617ab"}},{"code_sha256_prefix":"9d24359eae5d10c4","entry":"random_shear","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"image.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/image.py","link_basis":"harvester_set","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":"9d24359eae5d10c4"}},{"code_sha256_prefix":"ec1c67befd8a9ab3","entry":"random_shift","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"image.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/image.py","link_basis":"harvester_set","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":"ec1c67befd8a9ab3"}},{"code_sha256_prefix":"969bfd9c98f79209","entry":"read_cifar10","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"cifar10/cifar10_input.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/cifar10/cifar10_input.py","link_basis":"harvester_set","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":"969bfd9c98f79209"}},{"code_sha256_prefix":"07793677326f6721","entry":"svd_orthonormal","repo":"ducha-aiki/LSUVinit","repo_kind":"official","path":"tools/extra/lsuv_init.py","file_url":"https://github.com/ducha-aiki/LSUVinit/blob/HEAD/tools/extra/lsuv_init.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"07793677326f6721"}},{"code_sha256_prefix":"962d5f8133e2c7ce","entry":"svd_orthonormal","repo":"JonasWechsler/DeepLearningLab5","repo_kind":"listed","path":"lsuv_init.py","file_url":"https://github.com/JonasWechsler/DeepLearningLab5/blob/HEAD/lsuv_init.py","link_basis":"harvester_set","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":"962d5f8133e2c7ce"}},{"code_sha256_prefix":"69c8426f8140deac","entry":"svd_orthonormal","repo":"ducha-aiki/LSUV-pytorch","repo_kind":"listed","path":"LSUV.py","file_url":"https://github.com/ducha-aiki/LSUV-pytorch/blob/HEAD/LSUV.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"69c8426f8140deac"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}