📄 中文摘要
在知识系统中，两个智能体可能拥有相同的知识库和相同的问题，但却给出不同的答案。这种现象并不是幻觉或智能体质量的问题，而是一个被忽视的确定性问题。人们通常关注知识系统的完整性和准确性，但很少提及一个重要的要求：任何智能体在以任何顺序读取信息源后，必须得出相同的系统模型。这一要求在知识系统的设计和应用中至关重要。
你的知识，你的模型 — 第三部分：确定性并不等于准确性
出处: Your Knowledge, Your Model — Part 3: Determinism Is Not Accuracy
发布: 2026年3月31日
📄 English Summary
Your Knowledge, Your Model — Part 3: Determinism Is Not Accuracy
Two agents can possess the same knowledge base and face the same question yet provide different answers. This discrepancy is not due to hallucination or agent quality issues, but rather a determinism problem that is often overlooked. While completeness and accuracy are commonly emphasized in knowledge systems, there exists a crucial requirement that is seldom explicitly mentioned: any agent, regardless of the order in which they access the sources, must arrive at the same model of the system. This requirement is vital for the design and application of knowledge systems.