{"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/how-do-vision-transformers-work-1","title":"How Do Vision Transformers Work?","arxiv_id":"2202.06709","date":"2022-02-14","proceeding":"ICLR 2022 4","authors":["Namuk Park","Songkuk Kim"],"abstract":"The success of multi-head self-attentions (MSAs) for computer vision is now indisputable. However, little is known about how MSAs work. We present fundamental explanations to help better understand the nature of MSAs. In particular, we demonstrate the following properties of MSAs and Vision Transformers (ViTs): (1) MSAs improve not only accuracy but also generalization by flattening the loss landscapes. Such improvement is primarily attributable to their data specificity, not long-range dependency. On the other hand, ViTs suffer from non-convex losses. Large datasets and loss landscape smoothing methods alleviate this problem; (2) MSAs and Convs exhibit opposite behaviors. For example, MSAs are low-pass filters, but Convs are high-pass filters. Therefore, MSAs and Convs are complementary; (3) Multi-stage neural networks behave like a series connection of small individual models. In addition, MSAs at the end of a stage play a key role in prediction. Based on these insights, we propose AlterNet, a model in which Conv blocks at the end of a stage are replaced with MSA blocks. AlterNet outperforms CNNs not only in large data regimes but also in small data regimes. The code is available at https://github.com/xxxnell/how-do-vits-work.","url_abs":"https://arxiv.org/abs/2202.06709v4","url_pdf":"https://arxiv.org/pdf/2202.06709v4.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":"how-do-vision-transformers-work-1","repo_url":"https://github.com/xxxnell/how-do-vits-work","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"how-do-vision-transformers-work-1","repo_url":"https://github.com/Westlake-AI/A2MIM","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"how-do-vision-transformers-work-1","repo_url":"https://github.com/naver-ai/cl-vs-mim","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"how-do-vision-transformers-work-1","repo_url":"https://github.com/xxxnell/spatial-smoothing","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"specificity","task_name":"Specificity"}],"methods":[{"method_slug":"alternet","method_name":"AlterNet"}],"datasets_introduced":[],"methods_introduced":[{"slug":"alternet","name":"AlterNet","full_name":"AlterNet"}],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2202.06709","atlas_url":"https://app.syntology.ai/?focus=2202.06709","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":"2202.06709"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2202.06709\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2202.06709"},"call":"get_concepts_for_paper(arxiv_id=\"2202.06709\")"}],"by_repo":[{"repo":"xxxnell/how-do-vits-work","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["2105.12639"],"harvested_for_calls":[{"arxiv_id":"2105.12639","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2105.12639"},"call":"get_harvested_code_for_paper(arxiv_id=\"2105.12639\")"}]},{"repo":"Westlake-AI/A2MIM","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"naver-ai/cl-vs-mim","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"xxxnell/spatial-smoothing","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["2105.12639"],"harvested_for_calls":[{"arxiv_id":"2105.12639","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2105.12639"},"call":"get_harvested_code_for_paper(arxiv_id=\"2105.12639\")"}]}],"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":"a8afcfa51c72c96fe14317c114e4c304f3f44d63840d0151b7dabe19227dd504","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}