[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-579037-105":59,"doc-detail-579037-en":130},{"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":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","seeing-is-solving-mllms-reasoning-and-refusal-in-visual-math","Seeing Is Solving: MLLMs, Reasoning, and Refusal in Visual Math","","Many middle-school math problems depend on diagrams and figures, making them challenging for intelligent tutoring and for accessible learning when visuals are unavailable. This study evaluates six multimodal large language models on 376 Illustrative Mathematics items marked as image-role Required, testing with and without the figure under a shared protocol. Reasoning models perform notably better with images and produce adequate explanations, while without images models mostly refuse instead of guessing. Results inform gating and scaffold design, and the work releases code, prompts, and benchmark subsets for replication.",{"@graph":69,"@context":122},[70,84,105],{"@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/seeing-is-solving-mllms-reasoning-and-refusal-in-visual-math/579037/",{"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/seeing-is-solving-mllms-reasoning-and-refusal-in-visual-math/579037.png","ImageObject",300,407,{"name":92,"@type":93},"Noah","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-16","2026-10-15",true,{"@type":102,"interactionType":103,"userInteractionCount":24},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What makes the studied middle-school math problems “image-dependent”?","Question",{"text":112,"@type":113},"They rely on diagrams, graphs, or figures to specify task-critical quantities and relations that cannot be recovered from text alone without added assumptions.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How do the models behave when the required figure is missing?",{"text":117,"@type":113},"Models largely refuse rather than guess, with only rare correct-by-chance outcomes.",{"name":119,"@type":110,"acceptedAnswer":120},"What do the results suggest for intelligent tutoring and accessibility systems?",{"text":121,"@type":113},"Tutoring systems should gate scoring and learner updates on visual-evidence availability and use scaffolds that require explicit visual-evidence binding; accessibility systems should treat no-image refusals as missing-context signals and elicit the figure or a structured description.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},579037,1792209309,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":24,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"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},8796095462418,"https://ap-avatar.wpscdn.com/avatar/80000253c1241d02b47?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778826106357471780","Seeing Is Solving: MLLMs, Reasoning, and Refusal in Visual Math  \nEthan Croteau  \nWorcester Polytechnic Institute Worcester, MA[ecroteau@wpi.edu](ecroteau@wpi.edu)  \nNeil Heffernan  \nWorcester Polytechnic Institute Worcester, MA  \n[nth@wpi.edu](nth@wpi.edu)  \nMany middle–school math problems are image-dependent: the diagram or graph carries essential information. This matters for intelligent tutoring and accessibility, where systems must reason over figures and also decline responsibly when figures are missing. We evaluate six contemporary multimodal large language models (MLLMs)—three reasoning models and three non-reasoning models—on 376 Illustrative Mathematics (IM) items labeled as image-role Required (the figure contains task-critical information not recoverable from text alone without added assumptions) . Each model attempts every item three times with and without the figure under a shared prompt and scoring protocol. To reduce image-role label subjectivity, we classify items as not Required when they are solvable from text alone without additional assumptions.  \nWith images, the top-performing reasoning models achieve accuracy in the mid-50%, while nonreasoning models fall in the mid-30s to low-40s. Without images, models overwhelmingly refuse rather than guess, with only rare correct-by-chance answers. Models show moderate agreement on which items are solvable, and we release two benchmark subsets of items solved consistently across models. A qualitative audit of 83 items shows that visual misreading is the dominant failure mode for non-reasoning models, while reasoning models more often produce correct answers accompanied by adequate explanations.  \nThese results suggest tutoring systems should gate automated scoring and learner-model updates on visual-evidence availability and use scaffolds that require explicit visual-evidence binding before algebra. For accessibility, systems should treat no-image refusals as missing-context signals and elicit the figure or a structured description, enabling description-substitution experiments. We release code, prompts, and summary artifacts for replication. Code and data: [https://osf.io/ct7bg/](https://osf.io/ct7bg/)  \nKeywords: educational data mining, multimodal large language models, image-dependent math, refusal, reasoning, accessibility  \n1. INTRODUCTION  \nMany middle-school mathematics problems are image-dependent: they rely on diagrams, graphs, or other figures to specify quantities and relations that are not recoverable from text alone. Robust support for such items is critical for both intelligent tutoring systems (ITS) and accessibility. Decades of ITS research emphasize stepwise problem solving, mastery learning, and data-driven adaptation (Corbett and Anderson, 1994; Anderson et al., 1995 ; Woolf, 2008; VanLehn, 2011; Pelánek, 2017), and platforms such as ASSISTments instantiate these ideas at scale  \nin classroom-authentic settings (Heffernan and Heffernan, 2014; Roschelle et al., 2016 ; Selent et al., 2016) . In visual math, effective support requires binding labels and givens in the figure to symbols in the solution, checking scales and units on graphs, and making geometric relations explicit–practices aligned with diagram comprehension and multimedia learning principles (Larkin and Simon, 1987; Sweller et al., 1998 ; Ainsworth, 2006; Mayer, 2020) . Learners who cannot see or access images depend on alternative representations such as alt text, extended descriptions, audio, or tactile graphics. Accessibility guidelines for STEM materials stress that these alternatives must preserve task-relevant structure–naming entities, encoding geometric and quantitative relations, and clarifying scales and units–while managing cognitive load (DIAGRAM Center, 2015; W3C Accessibility Guidelines Working Group, 2024; W3C Web Accessibility Initiative, 2022; Braille Authority of North America (BANA), 2022; Marriott et al., 2021 ; Benetech, 2019; ASSETS 2025 Organizing Committee, 202","cbCaiuOecM53ZKeX","https://ap.wps.com/l/cbCaiuOecM53ZKeX","pdf",503410,42,"English","# Introduction\n# Related Work and Motivation\n## Image-Dependent Math for Tutoring and Accessibility\n## Key Evaluation Gaps for MLLMs\n# Study Overview and Experimental Setup\n## Image-Role Required Items and Labeling\n## With-Image vs Without-Image Testing","[{\"question\":\"What makes the studied middle-school math problems “image-dependent”?\",\"answer\":\"They rely on diagrams, graphs, or figures to specify task-critical quantities and relations that cannot be recovered from text alone without added assumptions.\"},{\"question\":\"How do the models behave when the required figure is missing?\",\"answer\":\"Models largely refuse rather than guess, with only rare correct-by-chance outcomes.\"},{\"question\":\"What do the results suggest for intelligent tutoring and accessibility systems?\",\"answer\":\"Tutoring systems should gate scoring and learner updates on visual-evidence availability and use scaffolds that require explicit visual-evidence binding; accessibility systems should treat no-image refusals as missing-context signals and elicit the figure or a structured description.\"}]","Seeing Is Solving: MLLMs, Reasoning, and Refusal in Visual Math | PDF",1792034885,106]