本文提出Chain of Mindset框架,通过动态选择不同认知模式提升推理效率,实验显示在多个基准测试中表现优异,准确率超越基线模型。
摘要:
人类解决问题并非单一认知模式的重复,而是整合多种认知模式。现有LLM推理方法固定单一认知模式,限制了智能发展。本文提出Chain of Mindset框架,分解推理为四种功能各异的认知模式:空间、收敛、发散和算法。Meta-Agent根据推理状态动态选择最优模式,双向上下文门控过滤跨模块信息流。实验表明,CoM在数学、代码生成、科学问答和空间推理等六个挑战性基准测试中达到SOTA性能,准确率超越基线模型4.96%和4.72%。代码已公开在https://github.com/QuantaAlpha/chain-of-mindset。
英文摘要:
Human problem-solving is never the repetition of a single mindset, by which we mean a distinct mode of cognitive processing. When tackling a specific task, we do not rely on a single mindset; instead, we integrate multiple mindsets within the single solution process. However, existing LLM reasoning methods fall into a common trap: they apply the same fixed mindset across all steps, overlooking that different stages of solving the same problem require fundamentally different mindsets. This single-minded assumption prevents models from reaching the next level of intelligence. To address this limitation, we propose Chain of Mindset (CoM), a training-free agentic framework that enables step-level adaptive mindset orchestration. CoM decomposes reasoning into four functionally heterogeneous mindsets: Spatial, Convergent, Divergent, and Algorithmic. A Meta-Agent dynamically selects the optimal mindset based on the evolving reasoning state, while a bidirectional Context Gate filters cross-module information flow to maintain effectiveness and efficiency. Experiments across six challenging benchmarks spanning mathematics, code generation, scientific QA, and spatial reasoning demonstrate that CoM achieves state-of-the-art performance, outperforming the strongest baseline by 4.96\% and 4.72\% in overall accuracy on Qwen3-VL-32B-Instruct and Gemini-2.0-Flash, while balancing reasoning efficiency. Our code is publicly available at \href{https://github.com/QuantaAlpha/chain-of-mindset}{https://github.com/QuantaAlpha/chain-of-mindset}.
Cognitive Inception: Agentic Reasoning against Visual Deceptions by Injecting Skepticism
CFAR, Agency for Science, Technology and Research (A*STAR), Singapore(科技研究局(A*STAR)认知智能研究所,新加坡);IHPC, Agency for Science, Technology and Research (A*STAR), Singapore(科技研究局(A*STAR)信息处理中心,新加坡);ROSE Lab, School of Electrical and Electronic Engineering, Nanyang Technological University(南洋理工大学电子与电气工程学院ROSE实验室)
CHARMS: A Cognitive Hierarchical Agent for Reasoning and Motion Stylization in Autonomous Driving
School of Mechanical and Electronic Engineering, Wuhan University of Technology(武汉理工大学机械与电子工程学院);Intelligent Transportation Systems Research Center, Wuhan University of Technology(武汉理工大学智能交通系统研究所);Department of Data and Systems Engineering, The University of Hong Kong(香港大学数据与系统工程系);School of Computer Science and Artificial Intelligence, Wuhan University of Technology(武汉理工大学计算机科学与人工智能学院);School of Mechanical Engineering, Southeast University(东南大学机械工程学院)
Latent Reward Steering: An Adaptive Inference-Time Framework that Implicitly Promotes Cognitive Behaviors in Reasoning LLMs
Rutgers University(罗格斯大学);South China Agricultural University(华南农业大学);Columbia University(哥伦比亚大学);Fenz.AI;QuantaAlpha;Adobe;Santa Clara University(圣克拉拉大学);City University of Hong Kong(香港城市大学)
Beyond Fast and Slow: Cognitive-Inspired Elastic Reasoning for Large Language Models
South China University of Technology(南方科技大学);Pazhou Laboratory(琶洲实验室);Peng Cheng Laboratory(鹏城实验室);Chongqing University of Posts and Telecommunications(重庆邮电大学)
MedCoG: Maximizing LLM Inference Density in Medical Reasoning via Meta-Cognitive Regulation
College of Computer Science, VCIP, DISSec Center, Nankai University, China 300350(南开大学计算机学院、VCIP、DISSec中心、中国300350);Generative AI Lab, College of Computing and Data Science(生成式人工智能实验室、计算与数据科学学院);Nanyang Technological University, Singapore 639798(南洋理工大学,新加坡639798)
Cognitive Pivot Points and Visual Anchoring: Unveiling and Rectifying Hallucinations in Multimodal Reasoning Models
Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学 龙城人工智能学院);South China Agricultural University(华南农业大学);South China Normal University(华南师范大学);Monash University(莫纳什大学);Jishou University(吉首大学);Hong Kong University of Science and Technology(香港科技大学);Tsinghua University(清华大学)
CAREBench: Evaluating LLMs' Emotion Understanding by Assessing Cognitive Appraisal Reasoning
Department of Informatics(信息学院);King’s College London(伦敦国王学院);Institute of Psychology(心理学研究所);University of Tartu(塔尔图大学);Department of Psychology(心理学系);Stanford University(斯坦福大学);The Alan Turing Institute(艾伦·图灵研究所)
Look on Demand: A Cognitive Scheduling Framework for Visual Evidence Acquisition in Multimodal Reasoning
Key Laboratory of Multimedia Trusted Perception(多媒体可信感知实验室);Efficient Computing, Ministry of Education of China, Xiamen University, 361005, P.R. China(教育部高效计算实验室,厦门大学,361005,中国);Sino-Russian ResearchCenter for Digital Economy(中俄数字经济研究中心);Institute of Artificial Intelligence, Xiamen University, China(人工智能研究院,厦门大学,中国);School of Informatics, Xiamen University, China(信息学院,厦门大学,中国);School of Information Engineering, Xiamen Ocean Vocational College, Xiamen 361102, China(信息工程学院,厦门海洋职业技术学院,厦门361102,中国)
PonsRAG: A Pons-Inspired RAG Bridging Cognitive Islands for Coordinated Long Narrative Reasoning
School of Future Technology, South China University of Technology(华南理工大学未来技术学院);TikTok Inc(字节跳动公司);School of Artificial Intelligence, Sun Yat-sen University(中山大学人工智能学院);School of Software Engineering, Sun Yat-sen University(中山大学软件工程学院)