[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82545-en":3,"doc-seo-82545-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82545,549758146520,"Patrick","https://ap-avatar.wpscdn.com/avatar/80002397d8c0411e94?_k=1775819394049821470",8,"Research & Report","Domain Arithmetic One-Shot VLA Adaptation under Environmental Shifts","Vision-Language-Action (VLA) models often lose performance when deployed under environmental shifts such as camera pose changes or transferring between similar robots (e.g., Panda to UR5e). Adapting to a target domain typically demands multiple task demonstrations, creating high data-collection costs. Domain ARiThmetic (DART) enables efficient one-shot adaptation by using weight vector arithmetic augmented with domain-specific information. DART isolates domain signals via subspace alignment and improves results across simulated and real-world visual and embodiment shifts.","arXiv :2607 .00666v 1 [ cs .RO] 1 Jul 2026  \nDomain Arithmetic: One-Shot VLA Adaptation under Environmental Shifts  \nTaewook Kang⋆, Taeheon Kim⋆, Donghyun Shin, and Jonghyun Choi†  \nSeoul National University  \n{tw.kang, thkim0305, dawnme, [jonghyunchoi}@snu.ac.kr](jonghyunchoi}@snu.ac.kr)  \nAbstract. Vision-Language-Action (VLA) models often fail to perform the same learned tasks under environmental shifts, such as changes in camera pose and shifts to a different but similar robot (e.g ., from Panda to UR5e) . Adapting these models to the shifted environment (i.e., target domain) often requires training on multiple demonstrations for each task, which are costly to collect. To reduce the burden of data curation and training, we propose an analogy-based method that adapts VLA models under environmental shifts through weight vector arithmetic with domainspecific information addition, named Domain ARiThmetic (DART) . Unlike prior approaches, DART requires collecting only a single demonstration, enabling efficient adaptation. To accurately isolate domain-specific information for addition, DART performs subspace alignment between singular components in weight vectors to filter out noisy components. In both simulated and real-world experiments, DART outperforms existing VLA adaptation methods in one-shot scenarios across diverse visual and embodiment shifts. Code is available at [https://github.com/snumprlab/dart](https://github.com/snumprlab/dart).  \nKeywords: Vision-Language-Action models · Environmental shifts · One-shot adaptation · Weight arithmetic  \n1 Introduction  \nVision-Language-Action (VLA) models trained on large-scale corpora show strong multi-task capabilities [3–5, 26, 27, 52, 80] . Despite their success within trained environments, i. e. , source domain, VLA models face challenges when deployed in new environments to perform learned tasks, a common real-world deployment scenario. These environmental shifts involve altered camera poses, distinct sensor calibrations, or embodiment modifications, leading to substantial performance degradation [13, 15, 33, 65, 74, 77, 79] . Thus, post-hoc adaptation remains essential to guarantee reliable execution in the shifted environment, i. e. , target domain. However, existing VLA adaptation approaches [1, 12, 13, 32, 63, 67] often require extensive expert demonstrations for every policy task in the target domain, resulting in severe deployment bottlenecks [42,60,72] . Also, fine-tuning on limited data often fails to generalize to unseen tasks [12, 67] .  \n⋆ These authors contributed equally.  \n† JC is with ECE, IPAI and ASRI in SNU and a corresponding author.  \n2 T. Kang et al.  \nFig. 1: One-shot VLA adaptation under environmental shifts. Environmental shifts can cause a source-trained VLA policy to fail in the novel target domain. (a) Fulldata fine-tuning adapts successfully but requires task-wise demonstrations, incurring expensive data collection costs. (b) One-shot fine-tuning is data-light but often fails to adapt across tasks. (c) Our one-shot adaptation extracts domain-specific directions from fine-tuned weights and adapts the policy via weight arithmetic.  \nFor practical policy deployment, extreme data efficiency for adaptation is essential in settings where collecting task-wise demonstrations at scale is typically infeasible, such as household environments. Thus, we aim for one-shot VLA adaptation where a policy adapts under environmental shifts using only a single demonstration of a single task. To enable this, we leverage the insight that a single demonstration can provide transferable domain knowledge. It allows a source-trained base VLA model to harness its learned task capabilities to solve the same tasks in the target domain, without relearning from scratch (Fig. 1) .  \nTo substantiate this idea, we analyze why one-shot fine-tuned models fail to adapt. Through subspace alignment analysis, we find that their parameter changes from a base model, i.e., update-vectors, a","cbCaibZBQA1pYeA4","https://ap.wps.com/l/cbCaibZBQA1pYeA4","pdf",5262939,1,43,"English","en",105,"# Introduction\n## One-shot VLA adaptation under environmental shifts\n## Motivation from update-vector subspaces\n## Proposed Domain ARiThmetic (DART)\n## Subspace filtering and subspace scaling","[{\"question\":\"Why do VLA models fail when moved to a target environment under environmental shifts?\",\"answer\":\"Environmental shifts such as altered camera poses, sensor calibration differences, or embodiment changes cause the source-trained policy to degrade in the target domain.\"},{\"question\":\"What problem does DART address in one-shot adaptation?\",\"answer\":\"DART targets the high cost of collecting multiple demonstrations per task by enabling adaptation with only a single demonstration of a single task in the target domain.\"},{\"question\":\"How does DART extract the domain information for weight arithmetic?\",\"answer\":\"DART subtracts source- and target-domain update-vectors to cancel shared task-specific directions, then uses subspace alignment-based filtering and scaling to remove misaligned or noisy components.\"}]",1784181451,108,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"domain-arithmetic-one-shot-vla-adaptation-under-environmental-shifts","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/domain-arithmetic-one-shot-vla-adaptation-under-environmental-shifts/82545/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why do VLA models fail when moved to a target environment under environmental shifts?","Question",{"text":75,"@type":76},"Environmental shifts such as altered camera poses, sensor calibration differences, or embodiment changes cause the source-trained policy to degrade in the target domain.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What problem does DART address in one-shot adaptation?",{"text":80,"@type":76},"DART targets the high cost of collecting multiple demonstrations per task by enabling adaptation with only a single demonstration of a single task in the target domain.",{"name":82,"@type":73,"acceptedAnswer":83},"How does DART extract the domain information for weight arithmetic?",{"text":84,"@type":76},"DART subtracts source- and target-domain update-vectors to cancel shared task-specific directions, then uses subspace alignment-based filtering and scaling to remove misaligned or noisy components.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]