用大模型知识构建轻量级世界模型，提升复杂场景下的预测与规划能力
CASSANDRA: Programmatic and Probabilistic Learning and Inference for Stochastic World Modeling
- 利用大模型生成代码建模确定性特征，指导概率图结构学习
- 在咖啡店和主题公园模拟器中，预测准确率显著优于基线方法
- 适合需要少样本、强因果理解的商业决策类系统研发者
构建世界模型对现实场景（如商业领域）的规划至关重要。由于这些领域语义丰富，可借助先验知识从有限数据中有效建模复杂动作效应与因果关系。本文提出CASSANDRA，一种神经符号世界建模方法，利用大语言模型（LLM）作为知识先验，构建轻量级转移模型以支持规划。该方法包含两个组件：(1) 由大模型合成的代码用于建模确定性特征；(2) 大模型引导的概率图模型结构学习，以捕捉随机变量间的因果关系。我们在两个场景中评估：(i) 小规模咖啡店模拟器，(ii) 复杂主题公园业务模拟器。结果表明，相比基线，CASSANDRA在转移预测与规划性能上均有显著提升。
原文摘要 · Abstract (English)
Building world models is essential for planning in real-world domains such as businesses. Since such domains have rich semantics, we can leverage world knowledge to effectively model complex action effects and causal relationships from limited data. In this work, we propose CASSANDRA, a neurosymbolic world modeling approach that leverages an LLM as a knowledge prior to construct lightweight transition models for planning. CASSANDRA integrates two components: (1) LLM-synthesized code to model deterministic features, and (2) LLM-guided structure learning of a probabilistic graphical model to capture causal relationships among stochastic variables. We evaluate CASSANDRA in (i) a small-scale coffee-shop simulator and (ii) a complex theme park business simulator, where we demonstrate significant improvements in transition prediction and planning over baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。