[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82169-en":3,"doc-seo-82169-105":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},82169,687197207057,"Sage","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","ARCANA Reflective Multi-Agent Program Synthesis Framework for ARC-AGI-2 Reasoning","ARCANA presents a collaborative multi-agent framework designed to solve ARC AGI 2 reasoning tasks under strict test-time and hardware constraints. Each task is decomposed into iterative perception, hypothesis generation, symbolic execution, and reflective refinement. A perceptual grounding agent builds object-centric scene graphs from grid inputs, while a latent program policy proposes diverse DSL programs. A symbolic executor verifies candidates on demonstrations, and a reflective agent generates failure-driven feedback for the next turn. Agents coordinate via a shared differentiable blackboard and a learned meta-controller, improving reasoning efficiency and solution quality on abstract transformation problems.","ARCANA: A Reflective Multi-Agent Program Synthesis Framework for ARC-AGI-2 Reasoning  \nKunbo Zhang * Columbia University  \nNew York, USA [kz2437@columbia.edu](kz2437@columbia.edu)  \nLei Fu  \nIndependent Researcher San Jose, USA[fuleiac@gmail.com](fuleiac@gmail.com)  \nZeYu Wang  \nUniversity of California, Los Angeles Los Angeles, USA [zeyuwang@ucla.edu](zeyuwang@ucla.edu)  \nZijing Liu  \nNortheastern University Boston, USA [liuzijing23@gmail.com](liuzijing23@gmail.com)  \nKejian Tong Independent Researcher  \nMukilteo, USA [tongcs2021@gmail.com](tongcs2021@gmail.com)  \narXiv :2607 .09059v 1 [ cs .AI] 10 Jul 2026  \nAbstract—We present ARCANA, a collaborative multi agent framework for solving ARC AGI 2 tasks under strict test time and hardware constraints. ARCANA decomposes each task into iterative perception, hypothesis generation, symbolic execution, and reflective refinement. A perceptual grounding agent builds object centric scene graphs from raw grids, a latent program policy proposes diverse DSL programs, a symbolic executor verifies candidates on demonstrations, and a reflective agent synthesizes failure driven feedback for the next turn. These agents communicate through a shared differentiable blackboard and are scheduled by a learned meta controller. The design combines structured program search with adaptive multi turn correction, improving reasoning efficiency and solution quality on challenging abstract transformation tasks.  \nIndex Terms—ARC AGI 2, multi agent reasoning, program synthesis, object centric learning, symbolic execution, test time adaptation  \nI. INTRODUCTION  \nAbstract reasoning on ARC AGI 2 remains difficult because successful solutions must infer compact transformation rules from only a few demonstrations, while handling variable grid sizes, object interactions, and severe ambiguity in the latent rule space. Large pretrained models have improved broad pattern recognition, yet few shot generalization on compositional grid transformations still demands precise search, explicit verification, and strong inductive bias for structure discovery [1] . Recent advances in reasoning have shown that intermediate deliberation can improve complex inference, but these gains do not directly solve ARC style tasks. Chain based prompting helps expose reasoning steps, yet it does not guarantee executable consistency across demonstrations [2] . Iterative self feedback further improves correction behavior, but pure text based refinement remains weak when the hypothesis space is symbolic, spatial, and tightly constrained by exact grid outputs [3] . To address this challenge, we introduce ARCANA, an adaptive collaborative architecture that turns each task into a multi turn reasoning episode. ARCANA separates object centric perception, latent program proposal, symbolic execution, and reflective refinement into specialized  \nagents connected by a shared blackboard. This design allows the system to generate diverse candidate programs, test them against demonstrations, diagnose failure patterns, and redirect search toward more promising regions of the program space. The result is a practical framework for abstract visual reasoning under realistic compute limits.  \nII. RELATED WORK  \nRecent work on structured reasoning has increasingly revisited object centric representations and modular rule manipulation as a foundation for compositional generalization. Slot based perceptual decomposition provides a differentiable route to discover entities and their attributes from raw inputs, which is especially useful when reasoning depends on object level transformations rather than dense pixels [4] . In parallel, neural systems that emulate production style rule application have shown how explicit modular structure can support more systematic computation than monolithic end to end predictors [5] . A second line of research emphasizes stronger reasoning behavior in neural models through explicit interaction between deliberation and action, or b","cbCaiiKgudsxTHd3","https://ap.wps.com/l/cbCaiiKgudsxTHd3","pdf",6492523,1,7,"English","en",105,"# Introduction\n# Related Work\n# Methodology","[{\"question\":\"How does ARCANA decompose an ARC AGI 2 task?\",\"answer\":\"ARCANA breaks each task into iterative perception, hypothesis generation, symbolic execution, and reflective refinement, forming a multi-turn reasoning episode.\"},{\"question\":\"What roles do the main agents play in ARCANA?\",\"answer\":\"A perceptual grounding agent constructs object-centric scene graphs, a latent program policy proposes DSL program candidates, a symbolic executor verifies them against demonstrations, and a reflective agent produces failure-driven feedback for subsequent turns.\"},{\"question\":\"How do agents coordinate and improve results during reasoning?\",\"answer\":\"Agents communicate through a shared differentiable blackboard and are scheduled by a learned meta controller, enabling structured program search with adaptive correction across turns.\"}]",1784178567,18,{"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},"arcana-reflective-multi-agent-program-synthesis-framework-for-arc-agi-2-reasoning","",{"@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/arcana-reflective-multi-agent-program-synthesis-framework-for-arc-agi-2-reasoning/82169/",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},"How does ARCANA decompose an ARC AGI 2 task?","Question",{"text":75,"@type":76},"ARCANA breaks each task into iterative perception, hypothesis generation, symbolic execution, and reflective refinement, forming a multi-turn reasoning episode.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What roles do the main agents play in ARCANA?",{"text":80,"@type":76},"A perceptual grounding agent constructs object-centric scene graphs, a latent program policy proposes DSL program candidates, a symbolic executor verifies them against demonstrations, and a reflective agent produces failure-driven feedback for subsequent turns.",{"name":82,"@type":73,"acceptedAnswer":83},"How do agents coordinate and improve results during reasoning?",{"text":84,"@type":76},"Agents communicate through a shared differentiable blackboard and are scheduled by a learned meta controller, enabling structured program search with adaptive correction across turns.","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},[93,97,101,105,110,115,119,122,127,130,134],{"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":21,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]