arXiv cs.AI· Qingshi Wang·· 13 小时前AI 评分31
超越大语言模型：用于心理建模、模拟、演化与可解释性的心理世界模型框架
Beyond Large Language Models: A Mental World Model Framework for Mental Modeling, Simulation, Evolution, and Interpretability
AI 导读
论文提出心理世界模型（MWM）及其实现多分支心理演化架构（MMEA），显式构建主观世界并模拟受历史经验塑造的连续心理演化。MMEA 在 DynToM 和 RecToM 上分别达到 96.48% 和 94.00% 准确率，超过领先 LLM 与专门 ToM 方法；在心理状态与行为预测、长序列预测、隐式心理状态推断上分别提升 28.7、39.3 和 16.6 个百分点。
正文 · 原文
Abstract:Understanding and simulating how subjective cognition and mental states emerge, evolve, and shape behavior is essential for advancing artificial intelligence toward deeper social cognition and general intelligence. Existing large language models (LLMs) and Theory of Mind (ToM) methods primarily infer mental states from linguistic information, lacking unified mechanisms for subjective world construction, cognitive-psychological interactions, dynamic evolution, and historical dependencies. This paper introduces the Mental World Model (MWM) and its implementation, the Multi-Branch Mental Evolution Architecture (MMEA), which explicitly constructs subjective worlds, models interactions between cognition and mental states, and simulates continuous mental evolution shaped by historical experiences. On DynToM and RecToM, MMEA achieves 96.48% and 94.00% accuracy, respectively, outperforming leading LLMs and specialized ToM methods. Additional experiments demonstrate improvements of 28.7, 39.3, and 16.6 percentage points in mental state and behavior prediction, long-sequence prediction, and implicit mental state inference, respectively. The results demonstrate that MMEA provides an interpretable framework for continuous mental modeling, enabling AI to move beyond surface-level semantic representations to infer individuals' underlying mental states from linguistic information and dynamically simulate their internal cognitive processes and evolution over time. MWM also provides a promising foundation for causal reasoning, personalized AI, sustained human-AI interaction, and future advances toward artificial general intelligence (AGI).
| Comments: | 12 pages, 10 figures |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.23722 [cs.AI] |
| (or arXiv:2610.23722v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23722 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Qingshi Wang [view email]
[v1]
Fri, 16 Oct 2026 16:42:11 UTC (1,524 KB)
来源：arXiv cs.AI · arxiv.org