A Visualized Chain of Thought System Based on the Symmetry Concept: Cracking the Dilemma of Complex Reasoning in Large Language Models

Abstract

Since 2023, large language models (LLMs) have achieved remarkable progress in natural language processing, but they still face dilemmas of unclear reasoning processes and low accuracy when tackling complex tasks such as logical reasoning and mathematical computation. The Chain of Thought (CoT) strategy has emerged as a viable solution to this problem, yet traditional CoT construction and management tools (e.g., LangChain) suffer from high abstraction and cumbersome operations. To address these issues, this study proposes a visualized CoT system integrated with the symmetry concept. The symmetry principle ensures consistency and coordination across all operational links, while the intuitive graphical interface and dynamic interactive functions simplify the entire process of CoT construction, adjustment, and management—effectively lowering the technical threshold for users. By presenting complex thinking structures in a visually and logically symmetrical form, the system enhances the orderliness and operability of complex reasoning processes. This work contributes to simplifying LLM reasoning workflows and promoting the application of CoT technology among non-technical users.