{"id":6132,"date":"2026-10-19T16:49:13","date_gmt":"2026-10-19T16:49:13","guid":{"rendered":"https:\/\/i10xnews.kinsta.cloud\/?p=6132"},"modified":"2026-10-19T17:19:17","modified_gmt":"2026-10-19T17:19:17","slug":"thinking-inertia-llms-keep-thinking-when-told-not-to","status":"publish","type":"post","link":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to","title":{"rendered":"Turn Off Thinking Mode and Models Still Reason 99.9% of the Time"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\">\u26a1 Quick Take<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>The finding:<\/strong> Even DeepSeek-V4-Flash (284B) shows visible explicit inference in 99.9% of open-ended responses when its thinking mode is disabled.<\/li><li><strong>Thinking inertia:<\/strong> Visible inference is organized more by the task&#8217;s answer space than by the thinking switch.<\/li><li><strong>The trade-off:<\/strong> Across five LLMs with fully disabled thinking, strict answer-only prompting raises mean Empty-Thinking Rate from near zero to roughly 40% while cutting mean accuracy by about 15 points.<\/li><\/ul>\n\n\n<p class=\"wp-block-paragraph\">Models increasingly ship with thinking and no-thinking modes. Researchers from Tsinghua, Oxford and Stanford argue that prior &#8220;no-thinking&#8221; proxies \u2014 a disabled switch or short traces \u2014 are unreliable. Disabled modes still emit extended reasoning, and long traces often contain filler.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1344\" height=\"768\" src=\"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-inertia.webp\" alt=\"Stat card: 99.9 percent residual inference with thinking disabled, about 40 percent empty-thinking rate under strict prompts with a 15-point accuracy cost, 98.8 percent judge agreement.\" class=\"wp-image-6131\" srcset=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-inertia.webp 1344w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-inertia-300x171.webp 300w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-inertia-1024x585.webp 1024w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-inertia-768x439.webp 768w\" sizes=\"auto, (max-width: 1344px) 100vw, 1344px\" \/><figcaption class=\"wp-element-caption\">i10X chart based on data from Lei et al. (2026), arXiv:2610.11765, main findings (CC0 1.0).<\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udd2c What the study shows<\/h2>\n\n\n<p class=\"wp-block-paragraph\">They normalize every response into a pre-answer trace and final answer, then score Empty-Thinking Rate (strict answer-only), Question\u2013Pre-answer Relevance, and an LLM-as-judge Explicit Inference Rate. Human studies support the metrics; annotators agree with the GPT-5.5 judge on explicit inference at 98.8% mean binary agreement. Six prompting interventions on six open and commercial LLMs.<\/p>\n\n\n<ul class=\"wp-block-list\"><li><strong>Switch \u2260 silence:<\/strong> Explicit no-think controls are not reliable no-thinking controls.<\/li><li><strong>Format matters:<\/strong> Fact-like and candidate-visible formats make answer-only responses easier than open generation for the same question.<\/li><li><strong>Not just difficulty:<\/strong> The pattern is not explained by benchmark difficulty or answer length alone.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u2705 Why it matters: controls that do not control<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Safety and cost work that assumes &#8220;no thinking&#8221; from a switch can silently still pay for reasoning tokens \u2014 and still get reasoning content.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83e\uddea Method in brief<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Three-level metric stack validated with human boundary audits and relevance judgments across six models and six intervention settings.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u2753 What remains open<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>Judge dependence:<\/strong> EIR uses an LLM judge (primary GPT-5.5).<\/li><li><strong>Prompt set:<\/strong> Results depend on the intervention wording.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\ud83e\udded Context<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Related reasoning-control work in <a href=\"\/news\/category\/studies\/reasoning\">Reasoning<\/a>.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udee0\ufe0f What it means for product and eval teams<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>Measure empty-thinking rate:<\/strong> Do not trust the mode flag.<\/li><li><strong>Expect an accuracy tax:<\/strong> Answer-only compliance costs points.<\/li><li><strong>Shape the answer space:<\/strong> Constrained formats suppress inertia better.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u270d\ufe0f About the analysis<\/h2>\n\n\n<p class=\"wp-block-paragraph\">This article is based on the full text of the preprint &#8220;Thinking Inertia: LLMs Keep Thinking When Told Not To&#8221; (arXiv:2610.11765v1, submitted October 8, 2026). All numbers are taken from the paper&#8217;s text, tables and figures; i10X did not rerun the experiments. The chart was redrawn by i10X from main findings of the paper, which is published under a CC0 1.0 license.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udd2d i10x Perspective<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Thinking Inertia is a useful name for a frustrating fact: models do not obey the off switch. If you need silence before the answer, you have to train and measure for it \u2014 not flip a boolean.<\/p>\n\n","protected":false},"excerpt":{"rendered":"<p>Thinking Inertia study: DeepSeek-V4-Flash keeps visible inference in 99.9% of open-ended answers with thinking disabled. Strict answer-only prompts raise Empty-Thinking Rate to ~40% but cut accuracy ~15 points.<\/p>\n","protected":false},"author":2,"featured_media":6282,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"paper_title":"Thinking Inertia: LLMs Keep Thinking When Told Not To","paper_url":"https:\/\/arxiv.org\/abs\/2610.11765","pdf_url":"https:\/\/arxiv.org\/pdf\/2610.11765v1","doi":"","arxiv_id":"2610.11765","authors":"Dianqiao Lei, Kevin Qinghong Lin, Pan Lu, Philip Torr, James Zou","institutions":"Tsinghua University; University of Oxford; Stanford University","paper_date":"2026-10-08","venue":"arXiv (cs.CL), version 1","peer_reviewed":false,"code_url":"","data_url":"","funding_coi":"See paper.","figure_license":"Chart drawn by i10X from main findings (CC0 1.0)","paper_license":"CC0 1.0","replication":"no","study_models":"Six LLMs incl. DeepSeek-V4-Flash; GPT-5.5 as EIR judge","study_sample":"Six prompting interventions; human validation of metrics","footnotes":"","_yoast_wpseo_primary_category":""},"categories":[56,52],"tags":[],"companies":[46,38],"products":[],"class_list":["post-6132","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-reasoning","category-studies","company-deepseek","company-open-ai"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.7 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Thinking Inertia: LLMs Keep Reasoning When Told Not To - i10X News<\/title>\n<meta name=\"description\" content=\"Study: disabled thinking modes still reason \u2014 99.9% for DeepSeek-V4-Flash. Answer-only prompts help (~40% ETR) but cost ~15 pp accuracy.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Thinking Inertia: LLMs Keep Reasoning When Told Not To - i10X News\" \/>\n<meta property=\"og:description\" content=\"Study: disabled thinking modes still reason \u2014 99.9% for DeepSeek-V4-Flash. Answer-only prompts help (~40% ETR) but cost ~15 pp accuracy.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to\" \/>\n<meta property=\"og:site_name\" content=\"i10X News\" \/>\n<meta property=\"article:published_time\" content=\"2026-10-19T16:49:13+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-10-19T17:19:17+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1344\" \/>\n\t<meta property=\"og:image:height\" content=\"768\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Christopher Ort\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:site\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Christopher Ort\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"2 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to\"},\"author\":{\"name\":\"Christopher Ort\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\"},\"headline\":\"Turn Off Thinking Mode and Models Still Reason 99.9% of the Time\",\"datePublished\":\"2026-10-19T16:49:13+00:00\",\"dateModified\":\"2026-10-19T17:19:17+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to\"},\"wordCount\":448,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp\",\"articleSection\":[\"Reasoning\",\"Studies\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to\",\"name\":\"Thinking Inertia: LLMs Keep Reasoning When Told Not To - i10X News\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp\",\"datePublished\":\"2026-10-19T16:49:13+00:00\",\"dateModified\":\"2026-10-19T17:19:17+00:00\",\"description\":\"Study: disabled thinking modes still reason \u2014 99.9% for DeepSeek-V4-Flash. Answer-only prompts help (~40% ETR) but cost ~15 pp accuracy.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp\",\"width\":1344,\"height\":768,\"caption\":\"Illustration: i10X\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/thinking-inertia-llms-keep-thinking-when-told-not-to#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/news.i10x.ai\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Turn Off Thinking Mode and Models Still Reason 99.9% of the Time\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#website\",\"url\":\"https:\\\/\\\/news.i10x.ai\\\/\",\"name\":\"i10X News\",\"description\":\"Stay updated with the latest news and announcements from i10X.\",\"publisher\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/news.i10x.ai\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\",\"name\":\"i10X\",\"url\":\"https:\\\/\\\/news.i10x.ai\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"contentUrl\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"width\":180,\"height\":180,\"caption\":\"i10X\"},\"image\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/i10x_ai\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/i10x-ai\",\"https:\\\/\\\/instagram.com\\\/i10x.ai\",\"https:\\\/\\\/www.youtube.com\\\/@i10X_ai_official\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\",\"name\":\"Christopher Ort\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"caption\":\"Christopher Ort\"},\"description\":\"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/author\\\/christopher-ort\",\"sameAs\":[\"https:\\\/\\\/i10x.ai\\\/blog\\\/author\\\/christopher-ort\"],\"worksFor\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\"},\"alumniOf\":{\"@type\":\"EducationalOrganization\",\"name\":\"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)\",\"url\":\"https:\\\/\\\/www.fau.eu\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Thinking Inertia: LLMs Keep Reasoning When Told Not To - i10X News","description":"Study: disabled thinking modes still reason \u2014 99.9% for DeepSeek-V4-Flash. Answer-only prompts help (~40% ETR) but cost ~15 pp accuracy.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to","og_locale":"en_US","og_type":"article","og_title":"Thinking Inertia: LLMs Keep Reasoning When Told Not To - i10X News","og_description":"Study: disabled thinking modes still reason \u2014 99.9% for DeepSeek-V4-Flash. Answer-only prompts help (~40% ETR) but cost ~15 pp accuracy.","og_url":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to","og_site_name":"i10X News","article_published_time":"2026-10-19T16:49:13+00:00","article_modified_time":"2026-10-19T17:19:17+00:00","og_image":[{"width":1344,"height":768,"url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp","type":"image\/webp"}],"author":"Christopher Ort","twitter_card":"summary_large_image","twitter_creator":"@i10x_ai","twitter_site":"@i10x_ai","twitter_misc":{"Written by":"Christopher Ort","Est. reading time":"2 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#article","isPartOf":{"@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to"},"author":{"name":"Christopher Ort","@id":"https:\/\/news.i10x.ai\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a"},"headline":"Turn Off Thinking Mode and Models Still Reason 99.9% of the Time","datePublished":"2026-10-19T16:49:13+00:00","dateModified":"2026-10-19T17:19:17+00:00","mainEntityOfPage":{"@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to"},"wordCount":448,"commentCount":0,"publisher":{"@id":"https:\/\/news.i10x.ai\/#organization"},"image":{"@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp","articleSection":["Reasoning","Studies"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#respond"]}]},{"@type":"WebPage","@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to","url":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to","name":"Thinking Inertia: LLMs Keep Reasoning When Told Not To - i10X News","isPartOf":{"@id":"https:\/\/news.i10x.ai\/#website"},"primaryImageOfPage":{"@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage"},"image":{"@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp","datePublished":"2026-10-19T16:49:13+00:00","dateModified":"2026-10-19T17:19:17+00:00","description":"Study: disabled thinking modes still reason \u2014 99.9% for DeepSeek-V4-Flash. Answer-only prompts help (~40% ETR) but cost ~15 pp accuracy.","breadcrumb":{"@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#primaryimage","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/thinking-inertia-llms-keep-thinking-when-told-not-to-title.webp","width":1344,"height":768,"caption":"Illustration: i10X"},{"@type":"BreadcrumbList","@id":"https:\/\/i10x.ai\/news\/thinking-inertia-llms-keep-thinking-when-told-not-to#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/news.i10x.ai\/"},{"@type":"ListItem","position":2,"name":"Turn Off Thinking Mode and Models Still Reason 99.9% of the Time"}]},{"@type":"WebSite","@id":"https:\/\/news.i10x.ai\/#website","url":"https:\/\/news.i10x.ai\/","name":"i10X News","description":"Stay updated with the latest news and announcements from i10X.","publisher":{"@id":"https:\/\/news.i10x.ai\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/news.i10x.ai\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/news.i10x.ai\/#organization","name":"i10X","url":"https:\/\/news.i10x.ai\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/news.i10x.ai\/#\/schema\/logo\/image\/","url":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","contentUrl":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","width":180,"height":180,"caption":"i10X"},"image":{"@id":"https:\/\/news.i10x.ai\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/i10x_ai","https:\/\/www.linkedin.com\/company\/i10x-ai","https:\/\/instagram.com\/i10x.ai","https:\/\/www.youtube.com\/@i10X_ai_official"]},{"@type":"Person","@id":"https:\/\/news.i10x.ai\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a","name":"Christopher Ort","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","caption":"Christopher Ort"},"description":"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.","url":"https:\/\/i10x.ai\/news\/author\/christopher-ort","sameAs":["https:\/\/i10x.ai\/blog\/author\/christopher-ort"],"worksFor":{"@id":"https:\/\/i10x.ai\/news\/#organization"},"alumniOf":{"@type":"EducationalOrganization","name":"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)","url":"https:\/\/www.fau.eu"}}]}},"_links":{"self":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6132","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/comments?post=6132"}],"version-history":[{"count":2,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6132\/revisions"}],"predecessor-version":[{"id":6283,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6132\/revisions\/6283"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media\/6282"}],"wp:attachment":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media?parent=6132"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/categories?post=6132"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/tags?post=6132"},{"taxonomy":"company","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/companies?post=6132"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/products?post=6132"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}