CoT-RAG: Integrating Chain of Thought and Retrieval-Augmented Generation to Enhance Reasoning in Large Language Models
Topics: AI (Deep Learning), Entity based search, Graph RAG, Knowledge Graph, LLMO / GEO, Retrieval Augmented Generation (RAG)
This paper introduces CoT-RAG, a novel framework that combines Chain of Thought reasoning with Retrieval-Augmented Generation to enhance the reasoning capabilities of Large Language Models. The framework features three key components: Knowledge Graph-driven CoT Generation, Learnable Knowledge Case-aware RAG, and Pseudo-Program Prompting Execution. Testing across nine public datasets showed significant accuracy improvements ranging from 4.0% to 23.0% compared to existing methods, with particularly strong performance on complex reasoning tasks.