[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-272675-105":59,"doc-detail-272675-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","arn-analogical-reasoning-on-narratives-v4-abstract","ARN - Analogical Reasoning on Narratives - v4 - Abstract","","Analogical reasoning supports transferable knowledge across domains, yet most NLP work focuses on word-level relational analogies rather than relational system analogies in narratives. This paper asks whether large language models can detect system analogies between narratives. It introduces a computational framework grounded in analogy theories, producing surface and system mappings, and a benchmark binary task, ARN, with far/near analogies and disanalogies. Results show near analogies are recognized, but far cases fail in zero-shot settings, while guided examples and chain-of-thought improve performance.",{"@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/arn-analogical-reasoning-on-narratives-v4-abstract/272675/",{"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/arn-analogical-reasoning-on-narratives-v4-abstract/272675.png","ImageObject",300,407,{"name":92,"@type":93},"Arica Lee","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-30","2026-09-15",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",18,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What problem does ARN aim to address?","Question",{"text":113,"@type":114},"ARN evaluates whether large language models can detect relational system analogies between narratives, extending beyond word-pair relational analogy work.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How is analogical reasoning modeled in the proposed framework?",{"text":118,"@type":114},"The framework operationalizes dominant analogy theories using narrative elements to create surface and system mappings, then leverages their interaction to build a benchmark.",{"name":120,"@type":111,"acceptedAnswer":121},"What are the main findings about LLM performance?",{"text":122,"@type":114},"LLMs largely recognize near analogies, but struggle with far analogies in zero-shot conditions, with GPT-4.0 scoring below random. Guidance via solved examples and chain-of-thought improves results, though best performance remains far from human level.","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},272675,1789466904,{"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":140,"language":141,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":67,"update_tm":130,"read_time":145,"preview_page_count":4},8796096645457,"https://ap-avatar.wpscdn.com/avatar/800003749518d68ffe3?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779345340919836971","ARN: Analogical Reasoning on Narratives Zhivar Sourati⋄ Filip Ilievski† Pia Sommerauer∗ Yifan Jiang⋄  \n⋄Information Sciences Institute, University of Southern California, Marina del Rey, CA, USA  \nDepartment of Computer Science, University of Southern California, Los Angeles, CA, USA †Department of Computer Science, Vrije Universiteit Amsterdam, Netherlands  \n∗ Computational Linguistics & Text Mining Lab, Vrije Universiteit Amsterdam, Netherlands {souratih,[yifjia}@isi.edu](yifjia}@isi.edu),{f.ilievski,[pia.sommerauer}@vu.nl](pia.sommerauer}@vu.nl)  \narXiv :2310 .00996v4 [ cs .CL] 3 Sep 2024  \nAbstract  \nAs a core cognitive skill that enables the transferability of information across domains, analogical reasoning has been extensively studied for both humans and computational models. However, while cognitive theories of analogy often focus on narratives and study the distinction between surface, relational, and system similarities, existing work in natural language processing has a narrower focus as far as relational analogies between word pairs. This gap brings a natural question: can state-of-theart large language models (LLMs) detect system analogies between narratives? To gain insight into this question and extend word-based relational analogies to relational system analogies, we devise a comprehensive computational framework that operationalizes dominant theories of analogy, using narrative elements to create surface and system mappings. Leveraging the interplay between these mappings, we create a binary task and benchmark for Analogical Reasoning on Narratives (ARN), covering four categories of far (cross-domain)/near (within-domain) analogies and disanalogies. We show that while all LLMs can largely recognize near analogies, even the largest ones struggle with far analogies in a zero-shot setting, with GPT4.0 scoring below random. Guiding the models through solved examples and Chain-of-Thought reasoning enhances their analogical reasoning ability. Yet, since even in the few-shot setting, the best model only performs halfway between random and humans, ARN opens exciting directions for computational analogical reasoners.  \n1 Introduction  \nAnalogical reasoning is a core cognitive skill unique to humans (Penn et al., 2008 ; Hofstadter,  \n2001), defined as the ability to perceive and utilize the similarities between situations or events based on (systems of) relations rather than surface similarities (Holyoak, 2012 ; Gentner et al., 2012) . The dichotomy between relational and surface similarity is illustrated in Figure 1 : while the narratives Q and N overlap in terms of characters, locations, and actions (surface similarity), they fail to create a system of relational correspondencesand thus are disanalogous as shown by their highlevel message. Meanwhile, Q and A are dissimilar on the surface while forming a coherent relational system of correspondences through the high-level message: no pain, no gain. Mappings based on systems of relations (system mappings) have priority over independent relations (Gentner et al., 1993) . Further, depending on whether the corresponding situations/events share the same domain or not, analogies and disanalogies can be qualified as near or far (Gentner, 1983) .  \nAnalogy enables creative inferences, explanations, and generalization of knowledge and has been used for scientific inventions (Dunbar and Klahr, 2012), solving problems (Gick and Holyoak, 1980), and policy-making (Houghton, 1998) . It has also been the subject of cognitive theories and studies about humans for common processes, such as retrieval of memories (Wharton et al., 1994) and problem-solving (Gick and Holyoak, 1980), mostly leveraging narratives as their experimental medium (e.g., Gentner and Toupin, 1986 ; Gentner et al., 1993 ; Wharton et al., 1994), given their multi-tiered nature and potential for abstraction. However, these studies focus on humans or models expecting well-structured inputs (Lu et al., 2022), are small-","cbCaipODcrk0C7VE","https://ap.wps.com/l/cbCaipODcrk0C7VE","pdf",1947411,23,"English","# Introduction\n## Analogy and relational vs. surface similarity\n## Motivation from cognitive studies\n## NLP analogies and existing benchmarks\n## Goal and gap addressed by ARN","[{\"question\":\"What problem does ARN aim to address?\",\"answer\":\"ARN evaluates whether large language models can detect relational system analogies between narratives, extending beyond word-pair relational analogy work.\"},{\"question\":\"How is analogical reasoning modeled in the proposed framework?\",\"answer\":\"The framework operationalizes dominant analogy theories using narrative elements to create surface and system mappings, then leverages their interaction to build a benchmark.\"},{\"question\":\"What are the main findings about LLM performance?\",\"answer\":\"LLMs largely recognize near analogies, but struggle with far analogies in zero-shot conditions, with GPT-4.0 scoring below random. Guidance via solved examples and chain-of-thought improves results, though best performance remains far from human level.\"}]","ARN - Analogical Reasoning on Narratives - v4 - Abstract | PDF",58]