[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-513842-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-513842-en":131},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","rase-retrieval-augmented-story-engine-control-in-game-using-ai","RASE: Retrieval Augmented Story Engine - Control in game using AI","","Recent advances in large language models (LLMs) enable coherent narrative generation, but purely generative story approaches often lose long-term coherence and factual consistency, producing disjointed or hallucinatory plotlines. The report presents a Retrieval-Augmented Story Engine (RASE) that couples a narrative-generating LLM with an external knowledge repository and retrieval mechanism. Retrieved context grounds each story segment on demand, reducing inconsistencies while preserving creativity. The document also analyzes RASE’s components and compares it with generative models, reinforcement-learning agents, and classical game master paradigms, situating retrieval augmentation within game AI.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/rase-retrieval-augmented-story-engine-control-in-game-using-ai/513842/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/rase-retrieval-augmented-story-engine-control-in-game-using-ai/513842.png","ImageObject",300,407,{"name":42,"@type":43},"Fahsai","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-11-15","2026-10-06",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",11,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What problem does RASE address in story generation?","Question",{"text":63,"@type":64},"It targets the tendency of purely generative approaches to break long-term coherence and factual consistency, leading to disjointed or hallucinatory storylines.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"How does RASE improve narrative quality?",{"text":68,"@type":64},"RASE grounds a narrative-generating LLM by coupling it with a retrieval mechanism and an external knowledge repository, fetching relevant context on demand for each story piece.",{"name":70,"@type":61,"acceptedAnswer":71},"Which approaches does the report compare against RASE?",{"text":72,"@type":64},"It compares RASE with purely generative narrative models, reinforcement learning-based storytelling agents, and classical game master or experience manager AIs.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},513842,1791369617,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,111,115,120,123,127],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":25,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":112,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":113,"slug":114},8,30,"research-report",{"id":116,"doc_module":4,"doc_module_name":25,"category_name":117,"show_sort_weight":118,"slug":119},9,"Religion & Spirituality",20,"religion-spirituality",{"id":118,"doc_module":4,"doc_module_name":25,"category_name":121,"show_sort_weight":118,"slug":122},"World Cup","world-cup",{"id":124,"doc_module":4,"doc_module_name":25,"category_name":125,"show_sort_weight":124,"slug":126},10,"Lifestyle","lifestyle",{"id":128,"doc_module":4,"doc_module_name":25,"category_name":129,"show_sort_weight":97,"slug":130},19,"General","general",{"code":4,"msg":82,"data":132},{"doc_id":79,"user_id":133,"nickname":42,"user_avatar":134,"doc_module":4,"category_id":112,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":140,"language":141,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":12,"update_tm":145,"read_time":146,"preview_page_count":33,"preview_urls":147},549768702563,"https://ap-avatar.wpscdn.com/avatar/8000c4aa63b76e948b?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786536092046926083","RASE: Retrieval Augmented Story Engine Narration  \nControl in game using AI  \nShishir Govinda M 1, N Chaitra2, Shiva Keerthi MP3, Rohit Mesta4  \n1,2,3,4Student, Dept. of CSE, Sir M Visvesvaraya Institute of Technology Bengaluru, India  \nAbstract—Recent advances in large language models (LLMs) have enabled AI systems to generate increasingly coherent and contextually rich narratives. However, purely generative approaches to story generation often struggle with maintaining long-term coherence and factual consistency, sometimes producing disjointed or hallucinatory storylines. To address these challenges, researchers have begun integrating retrieval augmented generation (RAG) techniques into storytelling systems. In a Retrieval-Augmented Story Engine (RASE), a narrativegenerating LLM is coupled with an external knowledge repository and retrieval mechanism, so that each piece of the story is grounded in relevant context fetched on demand. This framework leverages the strengths of both data-driven retrieval and generative modeling, aiming to reduce inconsistencies while preserving creativity. The following report examines RASE’s conceptual framework, core components, and implementation methodology, and situates it in the landscape of game AI and interactive narrative systems. We also compare RASE with alternative approaches – including purely generative narrative models, reinforcement learning-based storytelling agents, and classical game master or experience manager AIs – to clarify how retrieval augmentation builds upon and differs from these paradigms.  \nIndex Terms—RAG, SFT, Game, Narration, Machine Learning  \nI. INTRODUCTION  \nA. Developer Intention and Player Expectation  \nRole-playing games (RPGs) derive their allure from a tightly calibrated interplay between designer-specified mechanics and the emergent behaviors that players experience in situ, as elegantly characterized by the Mechanics–Dynamics–Aesthetics (MDA) framework and empirically demonstrated in recent anal- yses of player engagement. Players arrive with idiosyncratic objectivesand an innate proclivity to probe—and, at times, subvert—narrative constraints in pursuit of novel“dopamine hits,” a phenomenon that aligns closely with Flow theory’s stipulation of optimally balanced challenge  \nand skill. Yet despite these well-understood dynamics, the vast majority of RPGs remain anchored to static dialogue trees and linear scripting, which inherently stifle adaptive storytelling and constrain the generation of spontaneous, contextsensitive content. To address this limitation, we introduce a unified architecture—seamlessly embedding a vectorized in-game knowledge base, Retrieval-Augmented Generation (RAG), and SuperFine-Tuning (SFT) pipelines—directly within the game engine. Player-driven variability across divergent play through [1].  \nFig. 1. Developer Intention and Player Expectation  \nB. Player Stimulus  \nThe behavioral trace illustrated in figure below depicts live player engagement contours superimposed upon zones of boredom and anxiety, revealing a latent expanse for exploratory freedom that nonetheless remains rigidly bounded by preauthored content. In this model, even when players veer toward under- or over-stimulation, the prescribed action space rarely expands to accommodate emergent discoveries, resulting in a tacitly enforced “narrative inertia.” Such rigidity not only diminishes opportunities for serendipitous storytelling but also undermines the very sense of agency that underpins sustained immersion.  \nFig. 2. Developer’s Intent vs Player Curiosity  \nC. Allure of Modern RPGs  \nContemporary AAAA RPGs—exemplified by Red Dead Redemption 2, The Witcher 3, the Watch Dogs series, and Cyberpunk 2077—boast expansive open worlds teeming with NPCs, hidden interactions, and procedurally varied side-quests. Yet even these titans suffer from narrative fatigue: once players exhaust the limited repertoire of three to five dialogue variants per character (notably during","cbCaidP4awZBR0SU","https://ap.wps.com/l/cbCaidP4awZBR0SU","pdf",559711,15,"English","# Abstract\n# I. Introduction\n## A. Developer Intention and Player Expectation\n## B. Player Stimulus\n## C. Allure of Modern RPGs","[{\"question\":\"What problem does RASE address in story generation?\",\"answer\":\"It targets the tendency of purely generative approaches to break long-term coherence and factual consistency, leading to disjointed or hallucinatory storylines.\"},{\"question\":\"How does RASE improve narrative quality?\",\"answer\":\"RASE grounds a narrative-generating LLM by coupling it with a retrieval mechanism and an external knowledge repository, fetching relevant context on demand for each story piece.\"},{\"question\":\"Which approaches does the report compare against RASE?\",\"answer\":\"It compares RASE with purely generative narrative models, reinforcement learning-based storytelling agents, and classical game master or experience manager AIs.\"}]","RASE: Retrieval Augmented Story Engine - Control in game using AI | PDF",1791291914,38,[148,152,156,160],{"page_no":22,"url":149,"width":150,"height":151},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/04f1dcde7410a9462fb04e3c961150a5.webp",983,1272,{"page_no":26,"url":153,"width":154,"height":155},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/ff8c2c9396509245782e9175a22604ab.webp",982,1270,{"page_no":30,"url":157,"width":158,"height":159},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/c32cffb1371003f0ea5cec6653abd15f.webp",981,1269,{"page_no":33,"url":161,"width":162,"height":163},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/08437ae56312081dc7fb90a7d0eeca57.webp",980,1268]