The Illusion of Thinking: Models Don't Actually Reason

A paper posted to arXiv on 7 June 2025 and revised to v3 on 20 November 2025 by Parshin Shojaee and co-authors finds that Large Reasoning Models (LRMs) suffer a complete accuracy collapse beyond certain problem complexities, using controllable puzzle environments that allow precise manipulation of complexity while holding logical structure constant. The authors report a counterintuitive scaling limit: LRM reasoning effort rises with problem complexity up to a point, then declines even with remaining token budget, and they identify three regimes — standard LLMs outperform LRMs on low-complexity tasks, LRMs hold an advantage at medium complexity, and both collapse at high complexity. The paper also finds LRMs fail to use explicit algorithms for exact computation and reason inconsistently across scales, raising questions about their reasoning capabilities.

Computer Science Artificial Intelligence Submitted on 7 Jun 2025 v1 https://arxiv.org/abs/2506.06941v1 , last revised 20 Nov 2025 this version, v3 Title:The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity View PDF https://arxiv.org/pdf/2506.06941 HTML experimental https://arxiv.org/html/2506.06941v3 Abstract:Recent generations of language models have introduced Large Reasoning Models LRMs that generate detailed thinking processes before providing answers. While these models demonstrate improved performance on reasoning benchmarks, their fundamental capabilities, scaling properties, and limitations remain insufficiently understood. Current evaluations primarily focus on established math and coding benchmarks, emphasizing final answer accuracy. However, this evaluation paradigm often suffers from contamination and does not provide insights into the reasoning traces. In this work, we systematically investigate these gaps with the help of controllable puzzle environments that allow precise manipulation of complexity while maintaining consistent logical structures. This setup enables the analysis of not only final answers but also the internal reasoning traces, offering insights into how LRMs think. Through extensive experiments, we show that LRMs face a complete accuracy collapse beyond certain complexities. Moreover, they exhibit a counterintuitive scaling limit: their reasoning effort increases with problem complexity up to a point, then declines despite having remaining token budget. By comparing LRMs with their standard LLM counterparts under same inference compute, we identify three performance regimes: 1 low-complexity tasks where standard models outperform LRMs, 2 medium-complexity tasks where LRMs demonstrates advantage, and 3 high-complexity tasks where both models face complete collapse. We found that LRMs have limitations in exact computation: they fail to use explicit algorithms and reason inconsistently across scales. We also investigate the reasoning traces in more depth, studying the patterns of explored solutions and analyzing the models' computational behavior, shedding light on their strengths, limitations, and raising questions about their reasoning capabilities. Submission history From: Parshin Shojaee view email https://arxiv.org/show-email/cd7ce501/2506.06941 Sat, 7 Jun 2025 22:42:29 UTC 12,102 KB \ v1\ https://arxiv.org/abs/2506.06941v1 Fri, 18 Jul 2025 04:14:22 UTC 11,640 KB \ v2\ https://arxiv.org/abs/2506.06941v2 v3 Thu, 20 Nov 2025 00:19:24 UTC 11,960 KB Additional Features Current browse context: cs.AI References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .