KAG-Thinker: Interactive Thinking and Deep Reasoning in LLMs via Knowledge-Augmented Generation

TL;DR

KAG-Thinker introduces an interactive, multi-turn reasoning framework for LLMs that decomposes complex questions into sub-problems, improving logical coherence and knowledge integration in domain-specific question-answering tasks.

Contribution

The paper presents KAG-Thinker, a novel framework that enhances LLM reasoning with structured, multi-step problem decomposition and explicit knowledge retrieval mechanisms.

Findings

Improved logical coherence in LLM reasoning processes.

Effective multi-turn interaction for complex problem solving.

Enhanced knowledge integration from external sources.

Abstract

In this paper, we introduce KAG-Thinker, which upgrade KAG to a multi-turn interactive thinking and deep reasoning framework powered by a dedicated parameter-light large language model (LLM). Our approach constructs a structured thinking process for solving complex problems, enhancing the the logical coherence and contextual consistency of the reasoning process in question-answering (Q&A) tasks on domain-specific knowledge bases (KBs) within LLMs. Following the \textbf{Logical Form} guided retrieval and reasoning technology route of KAG, this framework first decomposes complex questions into independently solvable sub-problems (which are also referred to as logical forms) through \textbf{breadth decomposition}. Each such logical form is represented in two equivalent forms-natural language and logical function-and subsequently classified as either a Knowledge Retrieval or Reasoning Analysis task. Dependencies and parameter passing between these tasks are explicitly modeled via logical function interfaces. In the solving process, the Retrieval function performs retrieval tasks. It retrieves one-hop structured and unstructured information of specified knowledge unit. While the Math and Deduce functions are used to perform reasoning analysis tasks. Secondly, it is worth noting that, in the Knowledge Retrieval sub-problem tasks, LLMs and external knowledge sources are regarded as equivalent KBs. We use the \textbf{knowledge boundary} module to determine the optimal source using self-regulatory mechanisms such as confidence calibration and reflective reasoning, and use the \textbf{depth solving} module to enhance the comprehensiveness of knowledge acquisition...