[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84320-en":3,"doc-seo-84320-105":29,"detail-sidebar-cat-0-en-105":91},{"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},84320,687197207919,"Theodora","https://ap-avatar.wpscdn.com/avatar/a000253d6f5f7c60be?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779446848396160552",8,"Research & Report","MORES Mobile Reasoning-as-a-Service via Distributed LLM Inference-Time Scaling","Inference-time scaling enhances Large Language Models (LLMs) to improve reasoning strength without increasing model size. The work contrasts explicit reasoning that emits chain-of-thought tokens with implicit reasoning that updates hidden states in latent space, both of which create heavy compute and memory costs for edge deployment. MORES reframes reasoning as a wireless, edge-accessible service by splitting implicit updates between devices and servers. It jointly optimizes computation and communication using semantic MoE DRL, adapting recurrent steps and transmission pruning. Experiments report about 18% higher throughput versus SAC.","MORES: Mobile Reasoning-as-a-Service via Distributed LLM Inference-Time Scaling  \nGuanchen Liu, Hongyang Du∗ , Kaibin Huang  \n~~ ~~ ✦ ~~ ~~  \narXiv :2607 .08 1 16v 1 [ cs .NI] 9 Jul 2026  \nAbstract—Inference-time scaling has emerged as an effective approach for enhancing the capabilities of Large Language Models (LLMs), addressing the growing demand for stronger reasoning without increasing model size. This novel form of LLM scaling comprises two representative approaches: explicit reasoning, which generates intermediate chain-of-thought tokens during an explicit thinking phase, and implicit reasoning, which iteratively updates hidden states in the latent space without producing explicit outputs. Despite their effectiveness, both paradigms incur substantial computational and memory overhead, raising challenges for deployment on resource-constrained edge devices. To address these issues, we propose a Mobile Reasoning-as-aService (MORES) framework that treats reasoning as a computational service accessible to edge devices over wireless networks. Focusing on implicit reasoning, we leverage its recursive structure to partition hiddenstate updates between edge devices and servers, enabling cooperative inference that allows devices to access additional cloud computation on demand. To optimize long-term performance, we formulate a joint computation and communication scheduling problem and solve it using a semantic Mixture-of-Experts (MoE)-based Deep Reinforcement Learning (DRL) algorithm to address heterogeneity in wireless conditions and task demands. The agent adaptively allocates resources by adjusting the number of recurrent steps and the transmission pruning rate, while a semantic router enables high-speed gating for real-time expert selection. Experimental results show that the proposed method achieves an approximately 18% improvement in system throughput over the baseline Soft Actor-Critic (SAC) algorithm. Our code is available at [https://github.com/NICE-HKU/MORES](https://github.com/NICE-HKU/MORES).  \nIndex Terms—Large language models, reasoning, deep reinforcement learning, distributed computing, and wireless networks.  \n1 INTRODUCTION  \nLarge Language Models (LLMs) have scaled rapidly, with parameter counts increasing from 175 billion in OpenAI’s GPT-3 [1] to trillions in models such as GPT-5 [2], DeepSeekV4 [3], and Google’s Gemini 3 [4] . This parameter-driven growth followed predictable scaling laws [5], [6], where larger models outperformed smaller ones across diverse tasks. However, further expansion faces diminishing returns. Model training becomes prohibitively expensive, computational costs grow quadratically, and performance gains are increasingly constrained by data quality and availability. To overcome these limitations, researchers have introduced a new paradigm known as inference-time scaling [7] . Rather than expanding model parameters, this  \nG. Liu, H. Du, K. Huang are with the Department of Electrical and Computer Engineering, University of Hong Kong, Hong Kong SAR, China. (email: liugc@connect.hku.hk, duhy@hku.hk, huangkb@hku.hk)  \n\n| Device-only reasoning: limited by device capability, leading to degraded performance. |  |  |  |  |\n| --- | --- | --- | --- | --- |\n|  |  |  | No data exchange |  |\n\n\n| Cloud-only reasoning: limited by centralized processing, leading to server overload. |  |  |  |  |\n| --- | --- | --- | --- | --- |\n|  | \u003Cbr>What is the weather?\u003Cbr>\u003Cbr>Answer |  | Query\u003Cbr>\u003Cbr>\u003Cbr>It is a sunny day. |  |\n\n\n| MORES: leverages device computation and on-demand reasoning for superior performance. |  |  |  |  |\n| --- | --- | --- | --- | --- |\n|  |  |  |  |  |\n\nFig. 1: Comparison of reasoning modes between deviceonly, cloud-only, and the proposed device–cloud collaborative MORES architectures.  \napproach increases computation during inference to enhance reasoning capabilities. DeepSeek-R1 [8] represents a breakthrough in this direction by applying Deep Reinforcement Learning (DRL) to train LLMs tha","cbCain84vakmNHCv","https://ap.wps.com/l/cbCain84vakmNHCv","pdf",2949236,1,14,"English","en",105,"# Introduction\n# Mobile Reasoning-as-a-Service (MORES)\n## Implicit reasoning partitioning\n## Computation and communication scheduling\n# Experimental Results","[{\"question\":\"What problem does MORES address in deploying inference-time scaling on edge devices?\",\"answer\":\"Inference-time scaling improves reasoning but increases computation and memory overhead, making it difficult to run on resource-constrained edge devices and potentially overloading centralized servers. MORES targets this deployment challenge by enabling device–cloud collaborative reasoning.\"},{\"question\":\"How does MORES use implicit reasoning to enable distributed inference?\",\"answer\":\"MORES focuses on implicit reasoning and leverages its recursive structure to partition hidden-state updates between edge devices and servers, allowing devices to request additional cloud computation on demand.\"},{\"question\":\"How is the resource allocation policy optimized in MORES?\",\"answer\":\"MORES formulates a joint computation and communication scheduling problem and solves it with a semantic Mixture-of-Experts (MoE)-based Deep Reinforcement Learning (DRL) method. The agent adjusts recurrent steps and transmission pruning rate, while a semantic router performs fast expert selection.\"}]",1784194795,35,{"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":86,"head_meta":88,"extra_data":90,"updated_unix":27},"mores-mobile-reasoning-as-a-service-via-distributed-llm-inference-time-scaling","",{"@graph":35,"@context":85},[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/mores-mobile-reasoning-as-a-service-via-distributed-llm-inference-time-scaling/84320/",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",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does MORES address in deploying inference-time scaling on edge devices?","Question",{"text":75,"@type":76},"Inference-time scaling improves reasoning but increases computation and memory overhead, making it difficult to run on resource-constrained edge devices and potentially overloading centralized servers. MORES targets this deployment challenge by enabling device–cloud collaborative reasoning.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does MORES use implicit reasoning to enable distributed inference?",{"text":80,"@type":76},"MORES focuses on implicit reasoning and leverages its recursive structure to partition hidden-state updates between edge devices and servers, allowing devices to request additional cloud computation on demand.",{"name":82,"@type":73,"acceptedAnswer":83},"How is the resource allocation policy optimized in MORES?",{"text":84,"@type":76},"MORES formulates a joint computation and communication scheduling problem and solves it with a semantic Mixture-of-Experts (MoE)-based Deep Reinforcement Learning (DRL) method. The agent adjusts recurrent steps and transmission pruning rate, while a semantic router performs fast expert selection.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]