{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/co-training-2-l-submodels-for-visual","title":"Co-training $2^L$ Submodels for Visual Recognition","arxiv_id":"2212.04884","date":"2022-12-09","proceeding":null,"authors":["Hugo Touvron","Matthieu Cord","Maxime Oquab","Piotr Bojanowski","Jakob Verbeek","Hervé Jégou"],"abstract":"We introduce submodel co-training, a regularization method related to co-training, self-distillation and stochastic depth. Given a neural network to be trained, for each sample we implicitly instantiate two altered networks, ``submodels'', with stochastic depth: we activate only a subset of the layers. Each network serves as a soft teacher to the other, by providing a loss that complements the regular loss provided by the one-hot label. Our approach, dubbed cosub, uses a single set of weights, and does not involve a pre-trained external model or temporal averaging. Experimentally, we show that submodel co-training is effective to train backbones for recognition tasks such as image classification and semantic segmentation. Our approach is compatible with multiple architectures, including RegNet, ViT, PiT, XCiT, Swin and ConvNext. Our training strategy improves their results in comparable settings. For instance, a ViT-B pretrained with cosub on ImageNet-21k obtains 87.4% top-1 acc. @448 on ImageNet-val.","url_abs":"https://arxiv.org/abs/2212.04884v1","url_pdf":"https://arxiv.org/pdf/2212.04884v1.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":"co-training-2-l-submodels-for-visual","repo_url":"https://github.com/facebookresearch/deit","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViT-H@224 (cosub)","rank_in_archive_order":62,"of":1060,"metrics":{"Top 1 Accuracy":"88.0%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViT-L@224 (cosub)","rank_in_archive_order":82,"of":1060,"metrics":{"Top 1 Accuracy":"87.5%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"Swin-L@224 (cosub)","rank_in_archive_order":102,"of":1060,"metrics":{"Top 1 Accuracy":"87.1%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViT-B@224 (cosub)","rank_in_archive_order":155,"of":1060,"metrics":{"Top 1 Accuracy":"86.3%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"Swin-B@224 (cosub)","rank_in_archive_order":165,"of":1060,"metrics":{"Top 1 Accuracy":"86.2%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ConvNeXt-B@224 (cosub)","rank_in_archive_order":189,"of":1060,"metrics":{"Top 1 Accuracy":"85.8%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"PiT-B@224 (cosub)","rank_in_archive_order":190,"of":1060,"metrics":{"Top 1 Accuracy":"85.8%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViT-M@224 (cosub)","rank_in_archive_order":263,"of":1060,"metrics":{"Top 1 Accuracy":"85.0%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"RegnetY16GF@224 (cosub)","rank_in_archive_order":336,"of":1060,"metrics":{"Top 1 Accuracy":"84.2%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViT-S@224 (cosub)","rank_in_archive_order":466,"of":1060,"metrics":{"Top 1 Accuracy":"83.1%"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2212.04884","atlas_url":"https://app.syntology.ai/?focus=2212.04884","mcp":null,"developers":"https://syntology.ai/developers","held":true,"reason":"arxiv_id_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"agent_calls":[{"tool":"get_citation_path","arguments":{"paper_1":"2212.04884"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2212.04884\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2212.04884"},"call":"get_concepts_for_paper(arxiv_id=\"2212.04884\")"}],"by_repo":[{"repo":"facebookresearch/deit","kind":"named_in_paper","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":2,"harvested_for":["2112.13692","2403.14715"],"harvested_for_calls":[{"arxiv_id":"2112.13692","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2112.13692"},"call":"get_harvested_code_for_paper(arxiv_id=\"2112.13692\")"},{"arxiv_id":"2403.14715","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2403.14715"},"call":"get_harvested_code_for_paper(arxiv_id=\"2403.14715\")"}]}],"by_repo_harvested_for_is":"harvested_for: the arXiv ids (as Syntology's graph holds them) of every paper whose samples Syntology's graph links to this repository, in arXiv-id order, complete; the page names the first 3 and says how many more. A link only: it does not say the papers are the same work, that one copied the other, or that the code implements this paper's method; harvested_for_calls gives the get_harvested_code_for_paper call for each arXiv-shaped id among them","by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"f21d8c92fa7f7dfd11c921d5e389b0fea992f1f73e20a18fb107117804916f46","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}