Large language models have achieved remarkable capabilities across domains, yet mechanisms underlying sophisticated reasoning remain elusive. Recent reasoning models outperform comparable instruction-tuned models on complex cognitive tasks, attributed to extended computation through longer chains of thought. Here we show that enhanced reasoning emerges not from extended computation alone, but from simulating multi-agent-like interactions -- a society of thought -- which enables diversification and debate among internal cognitive perspectives characterized by distinct personality traits and domain expertise. Through quantitative analysis and mechanistic interpretability methods applied to reasoning traces, we find that reasoning models like DeepSeek-R1 and QwQ-32B exhibit much greater perspective diversity than instruction-tuned models, activating broader conflict between heterogeneous personality- and expertise-related features during reasoning. This multi-agent structure manifests in conversational behaviors, including question-answering, perspective shifts, and the reconciliation of conflicting views, and in socio-emotional roles that characterize sharp back-and-forth conversations, together accounting for the accuracy advantage in reasoning tasks. Controlled reinforcement learning experiments reveal that base models increase conversational behaviors when rewarded solely for reasoning accuracy, and fine-tuning models with conversational scaffolding accelerates reasoning improvement over base models. These findings indicate that the social organization of thought enables effective exploration of solution spaces. We suggest that reasoning models establish a computational parallel to collective intelligence in human groups, where diversity enables superior problem-solving when systematically structured, which suggests new opportunities for agent organization to harness the wisdom of crowds.

翻译：大型语言模型已在多个领域展现出卓越能力，但复杂推理背后的机制仍不明确。近期推理模型在复杂认知任务上表现优于同类指令微调模型，这通常归因于通过更长思维链实现的扩展计算。本文证明，增强的推理能力并非仅源于扩展计算，而是通过模拟多智能体式交互——即思想社会——实现的，这种机制促进了具有不同人格特质与领域专长的内部认知视角之间的多样化与辩论。通过对推理轨迹进行定量分析与机制可解释性方法研究，我们发现DeepSeek-R1与QwQ-32B等推理模型比指令微调模型展现出更显著的视角多样性，在推理过程中激活了更广泛的异质性人格特征与专业知识特征间的冲突。这种多智能体结构体现于对话行为（包括问答、视角转换与矛盾观点调和）以及表征激烈来回对话的社会情感角色中，共同构成了推理任务中的准确性优势。受控强化学习实验表明，当仅因推理准确性获得奖励时，基础模型会增强对话行为；而使用对话支架对模型进行微调，能比基础模型更快提升推理能力。这些发现表明，思想的社会化组织实现了对解空间的有效探索。我们认为推理模型建立了与人类群体集体智能的计算平行性——当多样性被系统化组织时，能够实现更优的问题解决，这为通过智能体组织利用群体智慧开辟了新的可能性。