Agentic Reasoning: A Streamlined Framework for Enhancing LLM Reasoning with Agentic Tools

TL;DR

This paper presents Agentic Reasoning, a framework that improves LLM reasoning by integrating external tools like web search and structured memory, achieving state-of-the-art results on complex tasks.

Contribution

The paper introduces the Mind-Map agent and an advanced Web-Search agent, enhancing reasoning coherence and search effectiveness in LLMs, surpassing previous approaches.

Findings

Achieved SOTA performance on DeepSeek-R1

Validated effectiveness of Mind-Map in reasoning tasks

Demonstrated superior web search capabilities

Abstract

We introduce Agentic Reasoning, a framework that enhances large language model (LLM) reasoning by integrating external tool-using agents. Agentic Reasoning dynamically leverages web search, code execution, and structured memory to address complex problems requiring deep research. A key innovation in our framework is the Mind-Map agent, which constructs a structured knowledge graph to store reasoning context and track logical relationships, ensuring coherence in long reasoning chains with extensive tool usage. Additionally, we conduct a comprehensive exploration of the Web-Search agent, leading to a highly effective search mechanism that surpasses all prior approaches. When deployed on DeepSeek-R1, our method achieves a new state-of-the-art (SOTA) among public models and delivers performance comparable to OpenAI Deep Research, the leading proprietary model in this domain. Extensive ablation studies validate the optimal selection of agentic tools and confirm the effectiveness of our Mind-Map and Web-Search agents in enhancing LLM reasoning. The code is at: https://github.com/theworldofagents/Agentic-Reasoning