[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-328500-105":59,"doc-detail-328500-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","rma-an-agentic-system-for-research-level-mathematical-problems","RMA - an Agentic System for Research-Level Mathematical Problems","","Research Math Agents (RMA) introduces an agentic framework for automated reasoning on research-level mathematical problems. Unlike approaches focused on competition math or fully formal theorem proving, RMA targets problems demanding long-horizon reasoning, literature grounding, and iterative proof refinement. It decomposes proof solving into specialized modules for analysis, literature search and understanding, comparison, knowledge-bank construction, and proof verification coordinated via initializer, proposer, and verifier agents using shared structured memory.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/rma-an-agentic-system-for-research-level-mathematical-problems/328500/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/rma-an-agentic-system-for-research-level-mathematical-problems/328500.png","ImageObject",300,407,{"name":92,"@type":93},"Violet","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-12","2026-09-21",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",13,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What is the main goal of Research Math Agents (RMA)?","Question",{"text":113,"@type":114},"RMA aims to automate reasoning for research-level mathematical problems that require long-horizon logic, literature grounding, and iterative refinement of proofs.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How does RMA coordinate different proof-solving capabilities?",{"text":118,"@type":114},"RMA decomposes the task into specialized modules—such as problem analysis, literature search and understanding, knowledge-bank construction, and proof verification—coordinated by initializer, proposer, and verifier agents via shared structured memory.",{"name":120,"@type":111,"acceptedAnswer":121},"How does RMA perform on the First Proof benchmark?",{"text":122,"@type":114},"RMA outperforms strong baselines, solving eight out of ten research problems and producing proofs that are more logically sound and readable based on comprehensive expert evaluation.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},328500,1790463839,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":56,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145,"preview_page_count":4},4398048950312,"https://ap-avatar.wpscdn.com/avatar/400002538284de19e3c?_k=1778320343897328908","arXiv :2605 .22875v1 [ cs .AI] 20 May 2026  \nRMA: an Agentic System for Research-Level Mathematical Problems  \nZelin Zhao Bo Yuan Jaemoo Choi Yongxin Chen  \nGeorgia Institute of Technology  \n{zzhao, byuan, jchoi, [yongxin.chen}@gatech.edu](yongxin.chen}@gatech.edu)  \nAbstract  \nWe present Research Math Agents (RMA), an agentic framework for automated reasoning on research-level mathematical problems. Unlike prior studies centered on competition mathematics or formal theorem proving, RMA targets research-level mathematical problems that require long-horizon reasoning, literature grounding, and iterative proof refinement. RMA decomposes research-level proof solving into specialized modules for problem analysis, literature search and understanding, fair comparison, knowledge-bank construction, and proof verification, all coordinated by initializer, proposer, and verifier agents through a shared structured memory.  \nWithin this unified framework, these agents operate in a multi-role, multi-round workflow, collaboratively generating, refining, and verifying candidate proofs through iterative feedback. We evaluate RMA on the First Proof benchmark, which consists often research-level problems contributed by expert mathematicians across diverse domains. Through comprehensive expert evaluation, RMA outperforms strong baselines on the First Proof benchmark, including GPT-5.2R and Aletheia, solving eight out of ten research problems and producing more logically sound and readable proofs. Our comprehensive ablation studies further show that performance gains arise from the interaction of structured reasoning modules, iterative refinement, and verifier-based feedback, rather than any single component. Our solutions and implementations will be made publicly available upon acceptance.  \n1 Introduction  \nMathematical reasoning has long served as a central benchmark for advanced artificial intelligence, as it requires precise symbolic manipulation, abstraction, and long-horizon logical reasoning [1] . Recent works have made substantial progress on competition-style benchmarks such as MATH [2] and even International Mathematical Olympiad (IMO) problems [3, 4] . Within these benchmark settings, a range of approaches have been explored, including formal mathematical proving systems [5, 6, 7] and LLM-assisted formal verification methods [8, 9, 10, 11, 12] . However, these benchmarks remain relatively well-scoped: problems are carefully selected to be solvable using known techniques and to admit concise, self-contained solutions, typically within one to two pages. In contrast, research-level mathematical problems pose a qualitatively different challenge [13, 14, 15] . They require the system to interpret specialized definitions, formalize assumptions, identify relevant prior results, explore effective proof strategies, construct intermediate lemmas, and carry out long chains of rigorous reasoning. In this scenario, formal systems are less well suited, as they typically assume fully specified problems and fixed formal contexts, whereas research-level mathematics requires engaging with relevant literature and developing proof strategies, often involving constructive techniques [16] .  \nRecently, there has been a surge of interest in agentic LLM systems that decompose complex tasks into specialized subtasks and solve them through planning, tool use, memory, critique, and iterative refinement [19, 20, 15, 21, 11, 22, 23, 24] . Such systems have shown promise in domains including software engineering [25], scientific discovery [11], and autonomous research workflows [14] . Despite this progress, research-level mathematics requires more than generic multi-step reasoning, and  \nResearch-level Problem:  \nMathematicians:  \nAletheia (DeepMind, 2026):  \nGPT-5 .2R (OpenAI, 2026):  \nOurs:  \n[Q6] For a graph G = (V, E) , let G! = (V, E S, S ) denote the graph with the same vertex set, but only the edges between vertices in S. Let L! be the Laplacian matrix of G ","cbCaihDklpV35zRD","https://ap.wps.com/l/cbCaihDklpV35zRD","pdf",474821,"English","# Introduction\n# Research-level Problem\n# Agents and Workflow\n# Evaluation on First Proof Benchmark\n# Ablation Studies\n# Case Study: Proof Derivation","[{\"question\":\"What is the main goal of Research Math Agents (RMA)?\",\"answer\":\"RMA aims to automate reasoning for research-level mathematical problems that require long-horizon logic, literature grounding, and iterative refinement of proofs.\"},{\"question\":\"How does RMA coordinate different proof-solving capabilities?\",\"answer\":\"RMA decomposes the task into specialized modules—such as problem analysis, literature search and understanding, knowledge-bank construction, and proof verification—coordinated by initializer, proposer, and verifier agents via shared structured memory.\"},{\"question\":\"How does RMA perform on the First Proof benchmark?\",\"answer\":\"RMA outperforms strong baselines, solving eight out of ten research problems and producing proofs that are more logically sound and readable based on comprehensive expert evaluation.\"}]","RMA - an Agentic System for Research-Level Mathematical Problems | PDF",1789979992,48]