[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-148120-105":59,"doc-detail-148120-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","scenegram-a-dataset-of-tangram-references-in-scene-context-research-paper","SCENEGRAM - A Dataset of Tangram References in Scene Context - Research paper","","Research on reference and naming shows that people can conceptualize the same object in diverse ways, while scene context shapes visual perception and conceptual expectations. This paper introduces SCENEGRAM, a dataset of human references to tangram shapes placed in different scene contexts, enabling systematic study of how context affects conceptualization. The work analyzes multimodal LLM-generated references and demonstrates that these models fail to capture the richness and variability of human conceptualizations.",{"@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/scenegram-a-dataset-of-tangram-references-in-scene-context-research-paper/148120/",{"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/scenegram-a-dataset-of-tangram-references-in-scene-context-research-paper/148120.png","ImageObject",300,407,{"name":92,"@type":93},"Riley","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-31","2026-08-26",true,{"@type":102,"interactionType":103,"userInteractionCount":39},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What is SCENEGRAM designed to study?","Question",{"text":112,"@type":113},"SCENEGRAM studies how scene context influences how people conceptualize and describe the same tangram shapes, using human reference data across different scene contexts.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"Why are tangrams useful in this research?",{"text":117,"@type":113},"Tangrams allow multiple valid conceptualizations (e.g., different labels for the same abstract shape), making them suitable for examining the connection between visual forms, scenes, and semantics.",{"name":119,"@type":110,"acceptedAnswer":120},"How do multimodal LLMs compare with human references?",{"text":121,"@type":113},"Experiments show that multimodal LLMs produce references that can resemble human patterns, but they do not capture the full richness and variability found in human conceptualizations and naming choices.","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},148120,1787775376,{"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":39,"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":129,"read_time":144,"preview_page_count":4},1374391975076,"https://ap-avatar.wpscdn.com/avatar/14000253ca4ec9f6853?x-image-process=image/resize,m_fixed,w_180,h_180&k=1783305029341752051","SCENEGRAM: Conceptualizing and Describing Tangrams in Scene Context  \nSimeon Junker and Sina Zarrieß Computational Linguistics, Department of Linguistics  \nBielefeld University, Germany  \n{simeon.junker,[sina.zarriess}@uni-bielefeld.de](sina.zarriess}@uni-bielefeld.de)  \nAbstract  \nResearch on reference and naming suggests that humans can come up with very different ways of conceptualizing and referring to the same object, e.g. the same abstract tangram shape can be a “crab”,“sink”, or “space ship”. Another common assumption in cognitive science is that scene context fundamentally shapes our visual perception of objects and conceptual expectations. This paper contributes SCENEGRAM, a dataset of human references to tangram shapes placed in different scene contexts, allowing for systematic analyses of the effect of scene context on conceptualization. Based on this data, we analyze references to tangram shapes generated by multimodal LLMs, showing that these models do not account for the richness and variability of conceptualizations found in human references.1  \n1 Introduction  \nReference to visual objects is an elementary component of language and situated interaction. Almost always, we can refer to something in many different ways, and the linguistic choices we make reflect the ways we categorize or conceptualize it. More often than not, we can have multiple conceptual perspectives (Clark, 1997) on the same things: For example, the same person could be referred to as a“human”,“woman”,“engineer”,“mom” or using their proper name, depending on what is deemed relevant, appropriate or useful in a given situation (e.g. Brown 1958 ; Graf et al. 2016) . Even fewer limitations exist for abstract shapes such as tangrams: In Figure 1, the same shape can be seen as representing e.g., a “crab”,“sink”, or “space ship”, showcasing the flexibility and richness of human conceptualization and interpretation at the intersection of visual and semantic processing. Despite recent progress in multimodal language modeling  \n1Data and code for this project are available at:  \n[github.com/clause-bielefeld/scenegram](github.com/clause-bielefeld/scenegram)  \n(Zhang et al., 2024), current systems show mixed results in reproducing human variation in object naming (Testoni et al., 2024) and figurative descriptions for abstract stimuli remain a major challenge in vision and language (V&L) research (Ji et al., 2022 ; Gul and Artzi, 2024) .  \nHow can we navigate this complex web of manyto-many relationships between visual stimuli and possible conceptualizations and descriptions? A common assumption in linguistics, cognitive science, and psychology is that our cognition is highly“tuned” to the everyday contexts and situations we interact in. A well-researched instance of this is our visual perception of objects which is known to be fundamentally shaped by high-level, conceptual expectations on the level of scenes (Biederman et al. 1982 ; Greene 2013 ; Võ 2021, among many others) . For example, at the beach, we would rather expect to see certain animals and plants than household items, whereas in a bathroom, it would be the other way around. It is well-researched that these processes facilitate, e.g., visual object recognition (e.g. Bar 2004), and it seems plausible that scene context also affects descriptions of objects which can be conceptualized in different ways. However, existing V&L datasets with real-world images make it difficult to study the same objects in different scenes: As objects often occur in typical contexts, the same regularities that are exploited for visual processing obstruct investigation with natural data. In this paper, we contribute (i) a dataset of human references to tangram shapes placed in different scene contexts, allowing for systematic analyses of the effect of scene context on the conceptualization of the same shape, and (ii) analysis of references to tangram shapes generated by multimodal LLMs, showing that these models do not ","cbCaiqMMgbBvhgbd","https://ap.wps.com/l/cbCaiqMMgbBvhgbd","pdf",2667709,17,"English","# Introduction\n# Background\n## Human scene understanding\n# Dataset and analysis\n## Human references to tangram shapes\n## Multimodal LLM reference analysis","[{\"question\":\"What is SCENEGRAM designed to study?\",\"answer\":\"SCENEGRAM studies how scene context influences how people conceptualize and describe the same tangram shapes, using human reference data across different scene contexts.\"},{\"question\":\"Why are tangrams useful in this research?\",\"answer\":\"Tangrams allow multiple valid conceptualizations (e.g., different labels for the same abstract shape), making them suitable for examining the connection between visual forms, scenes, and semantics.\"},{\"question\":\"How do multimodal LLMs compare with human references?\",\"answer\":\"Experiments show that multimodal LLMs produce references that can resemble human patterns, but they do not capture the full richness and variability found in human conceptualizations and naming choices.\"}]","SCENEGRAM - A Dataset of Tangram References in Scene Context - Research paper | PDF",43]