[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85131-en":3,"doc-seo-85131-105":29,"detail-sidebar-cat-0-en-105":83},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},85131,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Scaffolding the Strategist Architecture-Dependent Reasoning Interventions in Hotelling Spatial Markets","Structured reasoning interventions are tested for improving large language models’ strategic economic reasoning, focusing on whether intervention effects vary by model architecture. Using Hotelling’s linear city model, GPT-4.1-mini and GPT-5-mini are evaluated across an unscaffolded baseline and four interventions with eight deductive/abductive questions, multiple prompt framings, and repeated trials (720 judged responses). A significant scaffolding-type × architecture crossover emerges: commitment helps the standard model but harms the reasoning-optimized one, while principled separation reverses this. Adversarial stress degrades both models more severely for the reasoning model, and interventions address a declarative–procedural gap only for the reasoning model.","arXiv :2607 .09743v 1 [ cs .AI] 3 Jul 2026  \nScaffolding the Strategist: ArchitectureDependent Reasoning Interventions in Hotelling Spatial Markets  \nPratyush Singh  \nComputing + Mathematical Sciences California Institute of Technology Pasadena, CA 91125, USA [pksingh@caltech.edu](pksingh@caltech.edu)  \nAbstract  \nWe investigate whether structured reasoning interventions improve the strategic economic reasoning of large language models, and whether their effects depend on model architecture. Using Hotelling’s linear city model as a diagnostic vehicle, we evaluate GPT-4.1-mini (a standard instructionfollowing model) and GPT-5-mini (a reasoning-optimized model) under five conditions—an unscaffolded baseline and four reasoning interventions—across eight questions spanning deductive and abductive reasoning, three prompt framings, and three repetitions per condition, yielding 720 individually judged responses. We find a statistically significant crossover interaction between scaffolding type and model architecture (t(7) = 4 .79, p = 0 .002, d = 1 .69): commitment scaffolding improves the standard model (+0 .21) while degrading the reasoning model (−0 .63), and principled separation shows the opposite pattern (−0 .40 vs. +0 .31) . Both crossovers are individually significant (commitment: p = 0 .040; separation: p = 0 .002) and hold across all eight questions with 7/8 directional consistency. Adversarial stress-testing harms both models, with 2 .6 × greater degradation for the reasoning model (−1 .47 vs. −0 .57; p = 0 .038), and the damage correlates negatively with baseline difficulty (R2 = 0 .36, p = 0 .014) . We further document a persistent declarative–procedural gap in which both models identify correct strategies at rates far exceeding their ability to execute them; separation fully closes this gap for the reasoning model while no intervention helps the standard model.  \n1 Introduction  \nLarge language models have achieved remarkable performance on mathematical and logical benchmarks (OpenAI, 2023), yet their capacity for strategic economic reasoning—where agents must integrate domain knowledge, anticipate competitors, and translate understanding into concrete actions—remains poorly understood. Real-world decision-making requires multi-step strategic inference: reasoning backward from outcomes, maintaining consistency across choices, and distinguishing equilibrium predictions from naive intuitions.  \nA growing body of work has begun to address this gap by embedding LLM evaluation within game-theoretic frameworks. Studies using Cournot competition (Lin et al., 2024), repeated behavioral games (Akata et al., 2025), and decomposed strategic reasoning (Gandhi et al. , 2023) have consistently found that LLMs exhibit systematic biases invisible to standard benchmarks. GTBench (Duan et al., 2024) provides the most comprehensive game-theoretic LLM benchmark to date, spanning ten complete-and incomplete-information games, while EconArena (Guo et al., 2024) benchmarks economic reasoning using equilibrium deviation asa primary metric. However, these evaluations rely predominantly on discrete action spacesand simultaneous-move games, leaving continuous strategy spaces, multi-stage backward induction, and counterintuitive equilibria unexplored.  \nA critical dimension also remains unaddressed: how reasoning interventions interact with model architecture. The deployment of reasoning-optimized models with built-in chain-ofthought has raised questions about whether external scaffolding retains its value. Sprague et al. (2025) found diminishing returns from chain-of-thought prompting on models that already reason internally, and GTBench similarly reported that chain-of-thought does not always help in game-theoretic settings. Meanwhile, Zhang (2025) coined computational splitbrain syndrome to describe models that explain principles they cannot follow, a declarative– procedural gap that Gandhi et al. (2023) showed varies systematically across reas","cbCaidYapADUYAMc","https://ap.wps.com/l/cbCaidYapADUYAMc","pdf",1103649,1,26,"English","en",105,"# Abstract\n# Introduction\n# Related Work","[{\"question\":\"What happens under adversarial stress-testing, and does it affect both models equally?\",\"answer\":\"Adversarial stress-testing harms both models, with a larger degradation observed for the reasoning-optimized model than for the standard model. The damage also correlates negatively with baseline difficulty.\"}]",1784201288,66,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":78,"head_meta":80,"extra_data":82,"updated_unix":27},"scaffolding-the-strategist-architecture-dependent-reasoning-interventions-in-hotelling-spatial-markets","",{"@graph":35,"@context":77},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/scaffolding-the-strategist-architecture-dependent-reasoning-interventions-in-hotelling-spatial-markets/85131/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71],{"name":72,"@type":73,"acceptedAnswer":74},"What happens under adversarial stress-testing, and does it affect both models equally?","Question",{"text":75,"@type":76},"Adversarial stress-testing harms both models, with a larger degradation observed for the reasoning-optimized model than for the standard model. The damage also correlates negatively with baseline difficulty.","Answer","https://schema.org",{"og:url":51,"og:type":79,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":81,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":84},[85,89,93,97,102,107,112,115,120,123,127],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":86,"show_sort_weight":87,"slug":88},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":90,"show_sort_weight":91,"slug":92},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Exam",70,"exam",{"id":98,"doc_module":4,"doc_module_name":45,"category_name":99,"show_sort_weight":100,"slug":101},5,"Comic",60,"comic",{"id":103,"doc_module":4,"doc_module_name":45,"category_name":104,"show_sort_weight":105,"slug":106},6,"Technology",50,"technology",{"id":108,"doc_module":4,"doc_module_name":45,"category_name":109,"show_sort_weight":110,"slug":111},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":113,"slug":114},30,"research-report",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},9,"Religion & Spirituality",20,"religion-spirituality",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":121,"show_sort_weight":118,"slug":122},"World Cup","world-cup",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":124,"slug":126},10,"Lifestyle","lifestyle",{"id":128,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":98,"slug":130},19,"General","general"]