Can I read DeepRAG: Step-by-Step Retrieval for LLMs on EtoBox?
DeepRAG: Step-by-Step Retrieval for LLMs by yilin7yilin is a document available to read on EtoBox.
What is DeepRAG: Step-by-Step Retrieval for LLMs about?
The document introduces DeepRAG, a framework designed to enhance retrieval-augmented reasoning in large language models (LLMs) by modeling the process as a Markov Decision Process (MDP). DeepRAG improves retrieval efficiency and answer accuracy by dynamically determining whether to retrieve external knowledge or rely on parametric reasoning through iterative query decomposition. Experimental results demonstrate that DeepRAG significantly outperforms existing methods, achieving a 21.99% increase in accuracy
- Author
- yilin7yilin
- Language
- EN