當今最夯的議題，人工智能生成模型；人工智能推理模型之原理簡介如下：
A reasoning model refers to the study of how human thinking progresses from one idea to a related idea. It is an important aspect of problem-solving and involves the ability to reason over a multi-granular world. In computer science, a reasoning model is defined as an uncertain model on an AND/OR graph, where the efficiency of reasoning depends on the satisfaction degree of the homomorphism principle. It also involves the use of structures and operations to represent and approximate reasoning.
推理模型乃人類如何從一個想法到相關的想法思考過程之學習，對問題解決及涉及推理多粒度鎖世界是一個重要方向。於資訊科學領域，推理模型是建立於不確定之圖形模型上，於此有效率之模型憑藉同態原則符合程度；其亦涉及利用結構操作來表現相當的推理。