[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-327168-105":59,"doc-detail-327168-en":123},{"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":116,"head_meta":118,"extra_data":120,"updated_unix":122},105,"en","tudum-a-turkish-thinking-reasoning-pipeline-for-qwen35-27b","TÜDÜM: A Turkish-Thinking Reasoning Pipeline for Qwen3.5-27B","","This paper presents TÜDÜM (Türkçe Dü̧sünen Üretken Model), a project pipeline adapting a Qwen-family 27B thinking model toward Turkish reasoning. It targets a key distinction: not only producing Turkish final answers, but also making the explicit reasoning trace inside \u003Cthink>...\u003C/think> Turkish. The pipeline applies supervised fine-tuning on 15,991 Turkish reasoning examples with LoRA adapters, then uses GRPO-family reinforcement learning on a proxy-filtered Turkish mathematics environment. Results are mixed: SFT shortens responses and increases Turkish consistency yet lowers benchmark accuracy; RL recovers some math performance without surpassing the base model’s Macro-6 average. A step-50 model is released publicly.",{"@graph":69,"@context":115},[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/tudum-a-turkish-thinking-reasoning-pipeline-for-qwen35-27b/327168/",{"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/tudum-a-turkish-thinking-reasoning-pipeline-for-qwen35-27b/327168.png","ImageObject",300,407,{"name":92,"@type":93},"Violet","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-06","2026-09-21",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",18,{"@type":107,"mainEntity":108},"FAQPage",[109],{"name":110,"@type":111,"acceptedAnswer":112},"What overall performance trend is reported for SFT and RL?","Question",{"text":113,"@type":114},"SFT makes reasoning shorter and more consistently Turkish but reduces benchmark accuracy; RL recovers some mathematical performance yet does not uniformly improve all benchmarks or exceed the base model on the reported Macro-6 average.","Answer","https://schema.org",{"og:url":83,"og:type":117,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":119,"canonical":83},"index,follow",{"doc_id":121,"site_id":62},327168,1790365140,{"code":4,"msg":5,"data":124},{"doc_id":121,"user_id":125,"nickname":92,"user_avatar":126,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":127,"file_id":128,"file_url":129,"file_type":130,"file_size":131,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":29,"language":132,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":133,"faqs":134,"seo_title":135,"seo_description":67,"update_tm":136,"read_time":137,"preview_page_count":8,"preview_urls":138},4398048950312,"https://ap-avatar.wpscdn.com/avatar/400002538284de19e3c?_k=1778320343897328908","TÜDÜM: A Turkish-Thinking Reasoning Pipeline  \nfor Qwen3.5-27B  \nBaran Bingöl  \nDepartment of Artificial Intelligence and Data Engineering Faculty of Engineering  \nAnkara University  \nAnkara, Türkiye [22290007@ogrenci.ankara.edu.tr](22290007@ogrenci.ankara.edu.tr)  \nBahaeddin Türko˘glu  \nDepartment of Artificial Intelligence and Data Engineering Faculty of Engineering Ankara University  \nAnkara, Türkiye [turkoglub@ankara.edu.tr](turkoglub@ankara.edu.tr)  \narXiv :2607 .0 1927v 1 [ cs .CL] 2 Jul 2026  \nAbstract—This paper presents TÜDÜM (Türkçe Dü¸sünen Üretken Model), a project pipeline for adapting a Qwen-family 27B thinking model toward Turkish reasoning. The central problem is not only to answer Turkish prompts in Turkish, but to make the explicit reasoning trace itself Turkish. A thinking model may translate a Turkish prompt into an English-centered internal or visible scratchpad, solve the problem mostly in English, and only localize the final answer. TÜDÜM instead treats the generated \u003Cthink>...\u003C/think> block as a trainable behavior. The pipeline starts from the project base checkpoint unsloth/Qwen3.5-27B, applies supervised fine-tuning (SFT) on 15,991 Turkish reasoning examples using LoRA adapters, and then applies GRPO-family reinforcement learning on a proxyfiltered Turkish mathematics environment. The results are mixed. SFT made the model shorter and more consistently Turkish in its reasoning behavior, with large reductions in average response length and thinking exhaustion, but reduced benchmark accuracy. RL recovered some mathematical performance, especially AIME24 at the best early checkpoint, yet did not uniformly improve all benchmarks and did not exceed the base model on the reported Macro-6 average. The contribution is therefore best framed as a technically honest Turkish-thinking reasoning pipeline and evaluation, not as a claim of state-of-the-art Turkish reasoning. The released step-50 model is publicly available.  \n Model availability:  \n[https://huggingface.co/barandinho/qwen3.5-27b-tudum-dapo-50](https://huggingface.co/barandinho/qwen3.5-27b-tudum-dapo-50)  \nIndex Terms—Turkish reasoning, multilingual reasoning, supervised fine-tuning, reinforcement learning, GRPO  \nI. INTRODUCTION  \nReasoning models expose intermediate text before the final answer. In the TÜDÜM project this behavior is represented asa thinking block followed by a final response:  \n\u003Cthink>  \nintermediate reasoning trace  \n\u003C/think>  \nfinal answer  \nThis output format changes the multilingual problem. A model can receive a Turkish prompt and produce a Turkish final answer while still using English for most of the reasoning trace. This behavior may be acceptable for making the answer fit the local language. But this is not reasoning in Turkish. This difference is important for education, for checking how the answer was produced, and for later use in Turkish-language workflows. Because the user or evaluator does not only see  \nthe final answer. They also see the order of the mathematical, scientific, or coding decisions that led to that answer.  \nThe project base checkpoint is unsloth/Qwen3 .5-27B. Its model card lists Qwen/Qwen3 .5-27B as the base model, and the official Qwen3.5 model card states that Qwen3.5 models operate in thinking mode by default [1], [2] . For broader Qwen-family background, we also cite the Qwen3 technical report [3]. Prior work on chain-of-thought prompting showed that explicit intermediate steps can improve complex reasoning in large language models [4] . Multilingual chain-ofthought work further showed that reasoning can transfer across languages at sufficient scale [5] . However, newer analyses also suggest that multilingual models can use English-centered internal representations even when both input and output are non-English [6] . TÜDÜM is motivated by this gap: Turkish output is not sufficient evidence that the model reasons in Turkish.  \nThis paper summarizes the TÜDÜM pipeline and its measured behavior on th","cbCaikeeOIlbMB78","https://ap.wps.com/l/cbCaikeeOIlbMB78","pdf",369749,"English","# Abstract\n# I. Introduction\n# II. Method\n## A. Foundation Model and Thinking Format","[{\"question\":\"What overall performance trend is reported for SFT and RL?\",\"answer\":\"SFT makes reasoning shorter and more consistently Turkish but reduces benchmark accuracy; RL recovers some mathematical performance yet does not uniformly improve all benchmarks or exceed the base model on the reported Macro-6 average.\"}]","TÜDÜM: A Turkish-Thinking Reasoning Pipeline for Qwen3.5-27B | PDF",1789976584,15,[139],{"page_no":8,"url":140,"width":141,"height":142},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-08/54dc44c51baf36dd900795d1a421519a.webp",983,1390]