[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85943-en":3,"doc-seo-85943-105":30,"detail-sidebar-cat-0-en-105":92},{"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":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},85943,7971461740909,"Levi","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","SUREFlow: State-space Uncertainty-aware Residual Flow Matching for Robust Robot Manipulation","Generative vision-language-action policies improve robot manipulation but can become unstable under noise, partial observability, and stochastic starts. During long rollouts, small velocity prediction errors accumulate and reduce execution reliability. SUREFlow proposes a state-space uncertainty-aware residual flow matching framework with a Mamba backbone that jointly predicts action velocities and input-dependent residual uncertainty. It selectively refines unreliable action dimensions during inference without external feedback, maintaining efficiency. Results show 92.5% average success rate on LIBERO and about 49% on LIBERO-PRO with 179M parameters.","SUREFlow: State-space Uncertainty-aware REsidual Flow Matching for  \nRobust Robot Manipulation  \nMd Tanvir Islam, Sai Navaneet Peddapalli, Sangmoon Lee, Sangtae Ahn*  \narXiv :2607 . 10504v1 [ cs .RO] 11 Jul 2026  \nAbstract—Generative vision-language-action policies have advanced robot manipulation, but they often exhibit instability under noise, partial observability, and stochastic initial conditions. During extended rollouts, small velocity errors accumulate, degrading execution reliability. Existing diffusion and flow-based policies typically assume homoscedastic residualsand lack explicit uncertainty modeling within action generation, limiting robustness during iterative rollout. We propose SUREFlow, a state-space uncertainty-aware residual flow matching framework built on a Mamba backbone. The method jointly predicts action velocities and input-dependent residual uncertainty, enabling selective refinement of unreliable action dimensions without environment feedback while preserving computational efficiency. On LIBERO, SUREFlow achieves 92.5% average success rate (SR), outperforming the Mamba-based MaIL by 34.2% . On LIBERO-PRO, it attains around 49% SR using only 179M parameters, achieving performance comparable to large VLAs with 3-7B parameters. SUREFlow source code is available on: [https://github.com/tanvirnwu/SUREFlow](https://github.com/tanvirnwu/SUREFlow).  \nI. INTRODUCTION  \nLearning robot manipulation (RM) policies from demonstrations has become a dominant paradigm, where recent vision-language-action (VLA) models leverage multi-view visual observations and natural language conditioning to enable generalizable robotic skills across diverse tasks [1],[2], [3] . Despite this progress, many RM approaches still rely on direct action regression formulations [4], which yield deterministic predictions and are known to struggle when demonstrations exhibit multimodal behaviors, as commonly observed in real-world tasks [5] .  \nTo address these limitations, generative policy learning methods have recently gained attention. Diffusion-based and flow-based policies model the conditional distribution over actions rather than predicting a single output, enabling stochastic action generation and improved expressiveness, particularly in robotic manipulation settings [1], [6], [7] . Such methods have demonstrated improved robustness and flexibility compared to deterministic baselines. However, purely continuous generative policies typically operate at a single temporal scale and lack explicit temporal abstraction. As a result, stochasticity introduced during generation may accumulate over long action sequences, leading to unstable execution and inconsistent behaviors [5],[6] . Moreover, these approaches typically incur high computational costs due to  \nMd Tanvir Islam, Sai Navaneet Peddapalli, Sangmoon Lee, and Sangtae Ahn are with School of Electronic and Electrical Engineering, Kyungpook National University, Daegu 41566, Republic of Korea. (e-mails: tanvirnwu, navaneet, moony, [stahn @knu.ac.kr](stahn @knu.ac.kr))  \n*Corresponding author: Sangtae Ahn ([stahn@knu.ac.kr](stahn@knu.ac.kr))  \nFig. 1. Overview of URFlow, illustrating closed-loop refinement of uncertain action dimensions via internal residual updates during inference, without external feedback. The right panel shows representative LIBERO results, demonstrating consistent improvements over recent SOTA baselines.  \niterative sampling and often rely on transformer-based architectures, which substantially increase computational costs.  \nIn parallel, prior work has explored hierarchical and skillbased representations for RM, aiming to learn reusable skills directly from data rather than relying on hand-crafted primitives [8], [9] . However, most such approaches either separate skill learning from low-level control or depend on explicit planning modules, limiting seamless integration with end-to-end generative policies. Meanwhile, alternative sequence modeling architectures ","cbCaisRCUBfjTAVe","https://ap.wps.com/l/cbCaisRCUBfjTAVe","pdf",3834551,5,1,9,"English","en",105,"# Abstract\n# I. Introduction\n## Motivation and limitations of deterministic policies\n## Generative diffusion/flow policies and long-horizon instability\n## Efficiency constraints of transformer backbones\n## Gap in uncertainty handling for generative robot control\n## Proposed approach: SUREFlow\n## Contributions","[{\"question\":\"What problem does SUREFlow address in generative robot manipulation policies?\",\"answer\":\"SUREFlow targets instability during long-horizon rollouts caused by accumulated velocity errors under noise, partial observability, and stochastic initial conditions.\"},{\"question\":\"How does SUREFlow improve robustness during inference?\",\"answer\":\"It predicts input-dependent residual uncertainty and uses it to selectively refine unreliable action dimensions, reducing error accumulation without requiring external environment feedback.\"},{\"question\":\"What backbone and training formulation does SUREFlow use?\",\"answer\":\"SUREFlow builds a state-space uncertainty-aware residual flow matching policy on a Mamba backbone and formulates action generation as a continuous-time flow matching problem learned as a velocity field conditioned on observations and task embeddings.\"}]",1784207289,23,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"sureflow-state-space-uncertainty-aware-residual-flow-matching-for-robust-robot-manipulation","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/sureflow-state-space-uncertainty-aware-residual-flow-matching-for-robust-robot-manipulation/85943/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-07-26","2026-07-16",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What problem does SUREFlow address in generative robot manipulation policies?","Question",{"text":76,"@type":77},"SUREFlow targets instability during long-horizon rollouts caused by accumulated velocity errors under noise, partial observability, and stochastic initial conditions.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does SUREFlow improve robustness during inference?",{"text":81,"@type":77},"It predicts input-dependent residual uncertainty and uses it to selectively refine unreliable action dimensions, reducing error accumulation without requiring external environment feedback.",{"name":83,"@type":74,"acceptedAnswer":84},"What backbone and training formulation does SUREFlow use?",{"text":85,"@type":77},"SUREFlow builds a state-space uncertainty-aware residual flow matching policy on a Mamba backbone and formulates action generation as a continuous-time flow matching problem learned as a velocity field conditioned on observations and task embeddings.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,110,115,120,123,127,130,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":22,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":20,"slug":137},19,"General","general"]